Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
D 1077 OULU 2010 U N I V E R S I T Y O F O U L U P. O. B . 7 5 0 0 F I - 9 0 0 1 4 U N I V E R S I T Y O F O U L U F I N L A N D U N I V E R S I TAT I S S E R I E S SCIENTIAE RERUM NATURALIUM Professor Mikko Siponen HUMANIORA University Lecturer Elise Kärkkäinen TECHNICA Professor Hannu Heusala MEDICA Professor Olli Vuolteenaho SCIENTIAE RERUM SOCIALIUM ACTA TELERADIOLOGY— CHANGING RADIOLOGICAL SERVICE PROCESSES FROM LOCAL TO REGIONAL, INTERNATIONAL AND MOBILE ENVIRONMENT Senior Researcher Eila Estola SCRIPTA ACADEMICA Information officer Tiina Pistokoski OECONOMICA University Lecturer Seppo Eriksson EDITOR IN CHIEF Professor Olli Vuolteenaho PUBLICATIONS EDITOR Publications Editor Kirsti Nurkkala ISBN 978-951-42-6371-2 (Paperback) ISBN 978-951-42-6372-9 (PDF) ISSN 0355-3221 (Print) ISSN 1796-2234 (Online) U N I V E R S I T AT I S O U L U E N S I S Jarmo Reponen E D I T O R S Jarmo Reponen A B C D E F G O U L U E N S I S ACTA A C TA D 1077 UNIVERSITY OF OULU, FACULTY OF MEDICINE, INSTITUTE OF DIAGNOSTICS, DEPARTMENT OF DIAGNOSTIC RADIOLOGY; UNIVERSITY OF OULU, FACULTY OF MEDICINE, INSTITUTE OF BIOMEDICINE, DEPARTMENT OF MEDICAL TECHNOLOGY D MEDICA ACTA UNIVERSITATIS OULUENSIS D Medica 1077 JARMO REPONEN TELERADIOLOGY—CHANGING RADIOLOGICAL SERVICE PROCESSES FROM LOCAL TO REGIONAL, INTERNATIONAL AND MOBILE ENVIRONMENT Academic dissertation to be presented with the assent of the Faculty of Medicine of the University of Oulu for public defence in Auditorium 7 of Oulu University Hospital, on 10 December 2010, at 12 noon U N I VE R S I T Y O F O U L U , O U L U 2 0 1 0 Copyright © 2010 Acta Univ. Oul. D 1077, 2010 Supervised by Professor Osmo Tervonen Reviewed by Docent Tomi Kauppinen Professor Timo Paakkala ISBN 978-951-42-6371-2 (Paperback) ISBN 978-951-42-6372-9 (PDF) http://herkules.oulu.fi/isbn9789514263729/ ISSN 0355-3221 (Printed) ISSN 1796-2234 (Online) http://herkules.oulu.fi/issn03553221/ Cover Design Raimo Ahonen JUVENES PRINT TAMPERE 2010 Reponen, Jarmo, Teleradiology—changing radiological service processes from local to regional, international and mobile environment University of Oulu, Faculty of Medicine, Institute of Diagnostics, Department of Diagnostic Radiology, P.O.Box 5000, FI-90014 University of Oulu, Finland; University of Oulu, Faculty of Medicine, Institute of Biomedicine, Department of Medical Technology, P.O.Box 5000, FI90014 University of Oulu, Finland Acta Univ. Oul. D 1077, 2010 Oulu, Finland Abstract The possibilities of teleradiology to modify the radiological service process in a regional, international and mobile setting were investigated by building new types of technical connections and then by evaluating their feasibility. First a teleradiology link based on low-end technology was built for primary care and hospital settings. On evaluation, the total diagnostic agreement between the transmitted images and the original films was 98%. Then, a work practice-oriented approach was used to gain an understanding of the relationship between the emerging teleradiology work practice and the newly implemented technology. Ethnographically informed fieldwork and cooperative workshops were utilized. According to findings, articulation work that supports the key tasks is mostly conducted at the receiving site, and radiologists have to rely on much less information in image interpretation. The decisions made at the sending site influence the outcome. To study the idea of consultations between different countries, a connection utilizing the Internet was built between university hospitals in Oulu, Reykjavik and Tromsø. After 131 images, a suitable image compression ratio was selected. Image quality and transfer time of the 80 clinical case readings were found to be adequate for teleradiology. A wireless image consultation system for radiological sub-specialist consultations based on a portable computer and a mobile phone with secure access to the hospital network was set up and tested. The transmitted images of 68 patients were acceptable for final diagnosis in 72% of the cases. The wireless link saved the senior radiologist a hospital visit in 24% of the cases. A smartphone was then used to communicate computed tomography scans in a feasibility study of 21 patient cases of brain attacks. All transmitted image series were suitable for giving a preliminary consultation to the clinic, and in one case even a final report could be made. In a real life clinical setting of the study with neuroradiological and neurosurgical emergencies, two different smartphone platforms with electronic patient record integration were built in European research projects and evaluated with sets of 115 and 150 patient cases. They were good for final diagnosis in 38% and 40% of the cases, respectively. The concept was found to be ready for clinical use. Finally a survey was made showing the status and trends of the usage of eHealth applications in Finland. The results from all the public health care providers and a representative sample of private providers showed that in 2005, teleradiology services were used by 18/21 hospital districts and the usage of all eHealth applications has progressed throughout the entire health care delivery system. Teleradiology services have become an integrated part of eHealth. Keywords: computer-supported cooperative work, eHealth, radiography technology, telemedicine, teleradiology, wireless systems, work practice To my family 6 Acknowledgements This study was carried out at the Department of Diagnostic Radiology, Oulu University Hospital, in collaboration with the Department of Medical Technology, Institute of Biomedicine, University of Oulu, during the years 1995–2010. I owe my deepest gratitude to my supervisor, Professor Osmo Tervonen, M.D., Head of the Department of Diagnostic Radiology, for his encouragement during this work. He provided me with the opportunity and positive atmosphere to accomplish this work. He was always supportive in assuring that the work would be finished when the time was right. I wish to acknowledge Professor Emeritus Ilkka Suramo, M.D., the former Head of the Department of Diagnostic Radiology, who believed in my research topic and encouraged me to seek international collaborators and contacts early on in an emerging cross-discipline research field. I wish to express my gratitude to Professor Timo Jämsä, Ph.D. Head of Department of Medical Technology, who has enabled me to extend my studies into new research fields within his department. I also wish to express my particular gratitude to Docent Ilkka Winblad, M.D., my friend and closest co-worker in the past years. In addition to generously providing me with a lot of research knowledge, he has always kept reminding me how important it is to accomplish tasks once they are started. I wish to extend my warmest thanks to the official reviewers, Professor Timo Paakkala, M.D., and Docent Tomi Kauppinen, Ph.D., for their constructive criticism. Their efficient work and suggestions have markedly improved the outcome. This study would have not been possible without extensive teamwork. I wish to thank Antero Rahtu, M.Sc. (Engineering), for showing me the basics of teleradiology, as well as Thorgeir Palsson, M.Sc. (Engineering), Head of Clinical Engineering and Physics at Landspitalinn, Reykjavik, Iceland, and Torbjörn Sund, M.Sc. (Engineering), Telenor, Oslo, Norway, for showing me the possibilities of international data networks. I am grateful to Docent Eero Ilkko, M.D., and Docent Jaakko Niinimäki, M.D., who have always been open for new ideas and given the best of their knowledge and time for both scientific thinking and practical work in this study. My sincere thanks belong to Professor Seppo Lähde, M.D., and Juhani Laitinen, M.D., who besides taking part of this study, have been great clinical teachers in radiology. 7 All the other collaborators deserve my sincere thanks. From the Oulu team Professor Helena Karasti, Ph.D., Virpi Karhula, M.Sc., Docent Ari Karttunen, M.D., Docent Antero Koivula, Ph.D., Docent Timo Kumpulainen, M.D., Professor Kari Kuutti, Ph.D., Pekka Jartti, M.D., Ph.D., Lasse Jyrkinen, M.Sc. (Engineering), Eeva-Liisa Leisti, M.D., Ph.D., Professor Juhani Pyhtinen, M.D., Docent Markku Päivänsalo, M.D., and Docent Tarja Rissanen, M.D., Ph.D., have taken part in various new experiments without counting their working hours. I am grateful to Anna Vuolteenaho M.A., for the careful revision of the language of this thesis. I would also like to thank Professor Roberto Blanco, M.D., and Professor Miika Nieminen, Ph.D., for taking part in the follow-up group of this work. I also wish to thank Mrs Kaisa Punakivi, Mrs Marja-Liisa Raipola, Mrs Leila Salo and Ms Arja Väisänen for their kind assistance with practical matters. I express my gratitude to Virpi Honkala, M.D., chief physician at Raahe Hospital, who has been an inspiring clinical superior and done her best to guarantee possibilities to convert theories into practice. The board of the Finnish Society of Telemedicine and eHealth has made up an excellent atmosphere for discussing recent developments in eHealth. Many thanks to all the collaborators, especially to Sinikka Salo, DDS, Ph.D., and Principal Lecturer Anja Henner, Ph.D., for their support towards a scientific approach, and to Arto Holopainen, M.Sc. (Engineering), for technical support. A group of former photography enthusiasts, the Hemmos, has provided a yearly escape from formal activities. Special thanks are due to all of them, without forgetting Pertti Saloheimo, M.D., Ph.D., who also gave his editorial experience for the benefit of this work. Finally, I wish to express my warmest gratitude to my dear wife Jaana for her support throughout the years. Without her patiently taking care of daily responsibilities I would not have been able to concentrate on this work. My beloved children Tiina, Tuomo and Taneli have brought delight into my life and their music has carried me during the lonely evenings spent preparing these texts. I also give my dearest thanks to my parents, Laura and Paavo Reponen, for their support and guidance from early on. I am deeply sorry my mother was not able to see the finished work. The research was financially supported by European Commission R&D framework programmes IV (Health Telematics) and V (Information Society Technologies) and also by a grant from the Radiological Society of Finland, all of which are gratefully acknowledged. 8 Abbreviations 3G 4G AAPM ACR ANSI bit/s CCD CDMA CE CEN CR CSCW DCT DPI CT DICOM DF DM DR DSA EDGE EPR EU EVDO Funet FTP GPRS GSM HIS HL7 HSCSD HTTP ICT IEEE ISDN third-generation mobile networks fourth-generation mobile networks American Association of Physicists in Medicine American College of Radiology American National Standards Institute bits per second, used with prefixes like kbit/s, Mbit/s, Gbit/s charge coupled device code division multiple access conformité européenne, marking for European conformity Comité Européen de Normalisation computed radiography computer supported cooperative work discrete cosine transform dots per inch computed tomography, computerized tomography digital imaging and communication in medicine digital fluoroscopy digital mammography digital radiography digital subtraction angiography enhanced data rates for GSM Evolution electronic patient record European Union evolution data optimized Finnish University and Research Network file transfer protocol general packet radio service global system for mobile communications hospital information system health level 7 high-speed circuit-switched data hypertext transfer protocol information and communication technology Institute of Electrical and Electronics Engineers integrated services digital network 9 JIF JIRA JPEG bit/s LAN LCD LTE LZW Mhz MDD MR MRI MS-DOS NEMA NM NMT NORDUnet PACS PC PDA PPP QoS OSI RIS ROC SSL SMS TCP/IP TIFF TDMA UMTS US Unix VPN WAN WLAN WWW 10 JPEG interchange format Japan Industries Association of Radiological Systems Joint Photographers Expert Group bits per second local area network liquid crystal display long time evolution Lempel–Ziv–Welch megahertz Medical Devices Directive magnetic resonance magnetic resonance imaging MicroSoft Disk Operating System National Electrical Manufactures Association nuclear medicine nordic mobile telephone Nordic Infrastructure for Research & Education picture archiving and communication systems personal computer personal digital assistant point-to-point protocol quality of service open systems interconnect radiology information system receiver operating characteristic secure sockets layer short message service transmission control protocol /Internet protocol tagged image file format time division multiple access universal mobile telecommunications system ultrasound a computer operating system (mainly for efficient workstations) virtual private network wide area network wireless local area network World Wide Web List of original publications This thesis is based on the following original articles, which are referred to in the text by their Roman numerals: I II III IV V VI VII Reponen J, Lähde S, Tervonen O, Ilkko E, Rissanen T & Suramo I (1995) Low-cost Digital Teleradiology. Eur J Radiol 19: 226–231. Karasti H, Reponen J, Tervonen O & Kuutti K (1998) The teleradiology system and changes in work practices. Comput Methods Programs Biomed 57(1–2): 69–78. Reponen J, Palsson T, Kjartansson O, Ilkko E, Störmer J, Päivänsalo M, Pyhtinen J, Laitinen J, Eiriksson A & Brekkan A (1995) Nordic Telemedicine Consultation Network Using Internet. Proceedings of CAR ´95 Computer Assisted Radiology and Surgery. Berlin, Springer Verlag: 723–728. Reponen J, Ilkko E, Jyrkinen L, Karhula V, Tervonen O, Laitinen J, Leisti E-L, Koivula A & Suramo I (1998) Wireless radiological consultation with portable computers. J Telemed Telecare 4: 201–205. Reponen J, Ilkko E, Jyrkinen L, Tervonen O, Niinimaki J, Karhula V & Koivula A (2000) Initial experience with a wireless personal digital assistant as a teleradiology terminal for reporting emergency computerized tomography scans. J Telemed Telecare 6(1): 45–9. Reponen J, Niinimäki J, Kumpulainen T, Ilkko E, Karttunen A & Jartti P (2005) Mobile teleradiology with smartphone terminals as a part of a multimedia electronic patient record. Int Congr Ser 1281: 916–921. Reponen J, Winblad I & Hämäläinen P (2008) Current status of national eHealth and telemedicine development in Finland. Stud Health Technol Inform 134: 199–208. 11 12 Contents Abstract Acknowledgements 7 Abbreviations 9 List of original publications 11 Contents 13 1 Introduction 15 2 Review of the literature 17 2.1 Definition and development of teleradiology.......................................... 17 2.2 Change of paradigm: the digital image ................................................... 20 2.3 Required components of a teleradiology system..................................... 21 2.3.1 Image-sending station ................................................................... 21 2.3.2 Transmission network................................................................... 22 2.3.3 Image review station..................................................................... 23 2.3.4 Development of workstation and digitizer technology ................. 23 2.4 Recent communication technology developments .................................. 24 2.4.1 Wired connections ........................................................................ 24 2.4.2 Connection of networks; the Internet ........................................... 25 2.4.3 Mobile communication ................................................................. 26 2.5 DICOM-standard .................................................................................... 27 2.5.1 IHE integration profiles ................................................................ 28 2.6 Medical Devices Directive ...................................................................... 30 2.7 Image compression ................................................................................. 30 2.8 Data safety .............................................................................................. 32 2.9 Work process ........................................................................................... 33 2.10 Motivation of teleradiology..................................................................... 33 2.10.1 Effectiveness of teleradiology ...................................................... 33 2.10.2 Quality aspects ............................................................................. 35 2.11 Use of teleradiology ................................................................................ 36 2.11.1 Point to point consultations .......................................................... 36 2.11.2 Regional service concept .............................................................. 36 2.11.3 Mobile teleradiology .................................................................... 37 2.11.4 International teleradiology consultations...................................... 38 2.12 Concept of eHealth.................................................................................. 39 2.12.1 National initiatives for electronic patient documents ................... 40 2.12.2 EU cross border telemedicine initiatives ...................................... 40 13 2.12.3 Teleradiology as a part of eHealth services .................................. 41 3 Purpose of the study 43 4 Subjects and methods 45 4.1 Low-end technology teleradiology.......................................................... 45 4.2 Work practise analysis ............................................................................. 46 4.3 Internet teleradiology .............................................................................. 49 4.4 Mobile PC teleradiology ......................................................................... 50 4.5 Smartphone teleradiology ....................................................................... 54 4.5.1 Feasibility study............................................................................ 54 4.5.2 Clinical study ................................................................................ 55 4.6 Integration of teleradiology as a part of eHealth services ....................... 57 5 Results 59 5.1 Low-end technology teleradiology.......................................................... 59 5.2 Work practice analysis ............................................................................ 60 5.3 Internet teleradiology .............................................................................. 64 5.4 Mobile PC teleradiology ......................................................................... 66 5.5 Smartphone teleradiology ....................................................................... 67 5.5.1 Feasibility study............................................................................ 67 5.5.2 Clinical study ................................................................................ 68 5.6 Integration of teleradiology as a part of eHealth services ....................... 70 6 Discussion 75 6.1 Low-end technology teleradiology.......................................................... 76 6.2 Work practise analysis ............................................................................. 79 6.3 Internet teleradiology .............................................................................. 81 6.4 Mobile PC teleradiology ......................................................................... 83 6.5 Smartphone teleradiology ....................................................................... 85 6.5.1 Feasibility study............................................................................ 85 6.5.2 Clinical study ................................................................................ 89 6.6 Integration of teleradiology as a part of eHealth services ....................... 93 6.7 General discussion – from teleradiology to eHealth and beyond ............ 97 7 Summary and Conclusions 101 8 References 103 Original publications 119 14 1 Introduction Teleradiology services have long been performed with special solutions (Gitlin 1986, Batnitzky et al. 1990, Dwyer et al. 1991, Goldberg et al. 1993), but gradually connections came to be established with far less effort and cost than before (Dohrmann 1991, Yamamoto et al. 1996, Berry & Barry 1998, Engelmann et al. 2002). The advances in telecommunication and computer technology have given us widely standardized tools, which can be applied into clinical practice (Henri et al. 1997, Hussein et al. 2004). Teleradiology started as point-to-point connections using either telephone or dedicated data network connections (Thrall 2007a, Huang 2010). Two major advances forward have been the Internet (Caramella 1996) and mobile communication technology (Yamamoto 1995). They pose new challenges by linking experts quickly with patient information and breaking the barriers of time and physical hospital buildings. The Internet has even the potential to link international institutions to obtain secondary opinion (Kovalerchuk et al. 1998, Wagner 2002). Mobile phones have the potential of bringing a radiology workstation to a radiologist’s pocket (Yamamoto & Williams 2000). However, the technical feasibility of the Internet and mobile communication technology for clinical radiology practice needs to be assessed. New concepts have to be developed both in order to connect these tools to the hospital information systems (HIS) in an applicable and secure manner and to find the target medical areas and workflows that will benefit the most from these advances (Arenson et al. 2000, Benjamin et al. 2010). This means combining both engineering and medical knowledge in an effective way. The work process in a radiological department even within one institution is complicated by many successive steps (Siegel & Reiner 2002), and the utilization of the new telemedical workflow in this environment can therefore be problematic. The amount of teamwork needed to establish the final radiological report is extensive (Siegel & Reiner 2002). There are many challenges when we progress from an analogue to a digital environment (Reiner et al. 1996). These are related to technical, practical and quality aspects (Protopapas et al. 1996, Siegel et al. 1997, Soegner et al. 2003b). A work practice analysis can reveal aspects, otherwise hidden, that can be used to develop the teleradiology process chain (Abbot & Sarin 1994, Karasti 1997). When health care information processing turns digital, teleradiology becomes a natural part of eHealth (Harno 2004, Reponen 2008), which by definition 15 combines local and distant information and communication solutions in health care, such as electronic patient records (EPR), HIS, picture archiving and communication systems (PACS) and direct services for citizens under one umbrella (World Health Organization 2010). Before this study, the situation and the progression rate of digital image transfer in Finland has not been well known. Better knowledge of the present state also gives us tools to take the next logical steps and develop even international cross-border links (Ross et al. 2010) for consultations. The purpose of this study is to evaluate the use and implementation of teleradiology, more precisely the following: whether low-end technology could combine acceptable image quality and reasonable costs for primary care centres and local hospitals, whether there will be noticeable changes in the work processes when two different institutions start to work together, whether international teleconsultations using data networks could link together different institutions in a feasible manner, whether new ubiquitous communication tools, such as mobile data communication, are suitable for radiological purposes, whether smartphones could be used for image interpretations, and finally, to chart the usage status of teleradiology and digital radiology in the electronic health care (eHealth) domain in Finland in general. This study will discuss important research steps and achievements from traditional point-to-point teleradiology to mobile and internationally networked services over a time span of more than ten years. As some of the sub-topics in this study were among the earliest published work in this field, the results will draw a natural development path and summarize the newest trends in the rapidly developing field of teleradiology as part of the present electronic health care environment. 16 2 Review of the literature 2.1 Definition and development of teleradiology Teleradiology is one of the most established and most widely used forms of a larger entity called telemedicine (Wootton 1996, Roine et al. 2001, Ruotsalainen 2010). One of the early comprehensive definitions of telemedicine is that of Dripps et al. (1992), stating that telemedicine is the investigation, monitoring and management of patients and staff using systems which allow ready access to expert advice no matter where the patient is located. Telemedicine is actually an umbrella term that encompasses any medical activity involving an element of distance (Wootton 2001). A few years ago the term telemedicine began to be supplemented by the term telehealth (Watanabe et al. 1999, Krupinski et al. 2002), which was thought to be more “politically correct”, but since the year 2000 or thereabouts this too has been overtaken by even more fashionable terms, such as online health and e-health (or eHealth) (Wootton 2001). In a narrow definition, teleradiology has been defined as obtaining a specialist opinion by transmission of digital x-ray images to a radiologist elsewhere (often in a tertiary centre) (Wootton 2001). A wider definition by the Canadian Association of Radiologists (2008) says that teleradiology is the electronic transmission of diagnostic imaging studies from one location to another for the purposes of interpretation and/or consultation. It is mentioned that this definition covers both interfacility PACS networks as well as remote teleradiology. American College of Radiology (ACR) definition (2002) adds educational aspects by saying that teleradiology may allow even more timely interpretation of radiological images and give greater access to secondary consultations and improved continuing education. Images may be viewed simultaneously by users in different locations. According to ACR, when appropriately utilized, teleradiology can improve access to high-quality radiological interpretations and thus significantly improve patient care. The following is a short practical definition suitable for most purposes: Teleradiology is the ability to obtain images in one location, transmit them over a distance, and view them remotely for diagnostic or consultative purposes (Thrall 2007a). 17 The earliest mention of teleradiology dates back to 1929, when two dental radiographs were transmitted in the USA with the help of telegraph to a distant location. The image quality was cited to be good enough for dental fillings to be seen, according to an anonymous article published in the Journal of Dental Radiography and Photography (1929). This realized the hopes raised by an earlier anonymous article in published in the the New York Times (1907) describing the pioneering telephotograph work of professor Arthur Korn from Munich and predicting the coming wonders: “Lovers conversing at a great distance will behold each other as in the flesh. Doctors will examine patients’ tongues in another city, and the poor will enjoy visual trips wherever their fancy inclines.” The earliest generation of clinical teleradiology starting from the late 1950s was based on the transmission of analogue TV images with either close circuit or broadcast technology, using either real time or videotaped images (Jutras 1959, Murphy & Bird 1974, Murphy et al. 1970, Andrus et al. 1975a, Andrus et al. 1975b). However, the image quality was inadequate compared with the quality of film images (Page et al. 1981, Curtis et al. 1983). In addition, the working logistics were not optimal, as the images had to be sent one by one and the cost of a dedicated installation was high. Once these limitations were recognized, most of the pioneering projects of clinical teleradiology were terminated (Thrall 2007a). Then, in the early 1980s, attention turned to computer-based approaches to telemedicine, with a shift in interest from real-time television applications to “store-and-foreward” methods, in which data are collected in digital form at an initiating site and aggregated and stored for subsequent transmission to a receiving site (Gayler et al. 1981, Lufkin et al. 1983, Gitlin 1986, Thrall 2007a). Teleradiology systems became commercially available in the 1980s from a number of vendors, but in retrospect, they were very limited in terms of quality and scalability (Markivee & Chenoweth 1990, Thrall 2007a). So-called cameraon-a-stick systems enjoyed a brief period of popularity mostly for hospital-tohome applications to provide “after-hours” coverage (Bradley 2008). The approach entailed photographing or videographing selected hard-copy images for subsequent digitization and image transfer (Mun et al. 1995, Thrall 2007a). In the Nordic countries the first clinical teleradiology connection was established in Sweden in the early 1980s. A digital system for transmission of images over telephone lines using a 256 × 256 and 6 bit deep matrix from a video camera providing communication between a county hospital and a university hospital was presented. Transmission of computed tomography (CT) images was performed without significant degradation of image quality while deterioration of 18 image quality was noticed when conventional films were transmitted. The consultations gave clinically valuable information to the transmitting radiologist in approximately 50 percent of the cases. (Nilsson et al. 1986.) However, the limited image quality and relatively slow image transmission prevented the diffusion of this method (Olsson 1991). In Finland, distance teleradiology using the national television broadcasting network was tested briefly in 1969 (Soila 1970), but the results were only published in a non-scientific newspaper and magazine articles (Reponen 2004). Clinical teleradiology started when a videocamera-based photophone device with a 529 × 440 pixel and 8 bit deep matrix was taken into use between Jyväskylä central hospital and Kuopio University Hospital in 1989. CT images of neurosurgical emergencies were sent; each scanned image was transmitted over telephone lines in 30 seconds. According to users, the equipment saved its purchasing costs because transportation of patients could be avoided. (Vuorela & Puranen 1994.) More capable teleradiology, with film scanners and Unix computer workstations with a 1024 × 1024 and 2048 × 2048 pixel matrix that were still expensive at that time, was launched in Finland in the early 1990s (Viitanen et al. 1992) and evaluation results showed acceptable quality for chest images (Kormano et al. 1994, Korsoff et al. 1995). Lower resolution microcomputer-based teleradiology with a 512 × 512 pixel and 8-bit image matrix was also evaluated and found sufficient for consultation with most conventional radiographs in daily practice (Paakkala et al. 1991). In this study, images from 372 conventional roentgen examinations were digitized and transmitted via a 64 Kbits/s telephone line from a rural health centre to a university hospital, where they were interpreted by two radiologists. Slight deterioration of image quality was noticed, although the images were non-diagnostic only in a few cases. According to the authors, a system based on a 1024 × 1024 pixel matrix is desirable, however. Modern teleradiology thus began to develop after the introduction of digitized images. The early digitized systems captured video images with a 256 × 256 to 512 × 512 pixel matrix and a 6- to 8-bit dynamic range, while more sophisticated laser scanners enabled resolution of up to 4096 × 4096 pixel and 12 bit- grey scale (Jelaso et al. 1978, Goldberg et al. 1993). However, the approaches were cumbersome, and images were still handled one at a time (Thrall 2007a). Other obstacles for widespread use of teleradiology were the relatively low performance and high cost of computer systems available (Batnitzky et al. 1990), high costs of data transmission (Dwyer et al. 1991), and lack of practical and 19 affordable digital image handling systems including high-resolution workstations at originating and receiving sites (Thrall 2007a). 2.2 Change of paradigm: the digital image One of the key enabling factors for teleradiology has been the change in the overall paradigm of how radiological images are made and used. Since the invention of new X-rays (Röntgen 1895), the imaging process was for nearly a hundred years based on photographic chemistry and physical images. These images had to then be copied or transmitted via courier if their presence at another local was needed. New imaging modalities, like tomographic nuclear medicine (NM) imaging with computerized image display and storage in the 1960s and CT in 1971, were digital from the beginning and paved way to a digital medical image, which consisted of computer-made measurement results and had only a representation as a physical image (Langan & Wagner 1969, Thrall 2005, Beckmann 2006). Magnetic resonance imaging (MRI or MR), digital subtraction angiography (DSA), digital fluoroscopy (DF), ultrasound (US) with digital image capture followed soon after, and finally computed radiography (CR), digital radiography (DR) and digital mammography (DM) signalled the start of the digital age, making the computerized image commonplace (Kiuru et al. 1991b, Thrall 2005). First, these digital modalities revolutionalized image acquisition, resulting gradually in digital storage and transmission of images in everyday practice. PACS took care of the images (Huang et al. 1988, Taira et al. 1988, Kiuru et al. 1991a, Viitanen et al. 1991), while radiology information systems (RIS) managed the radiological workflow and information exchange with the other components in the HIS (Kotter & Langer 1998, Kinsey et al. 2000). As radiologists’ working environment has moved away from a lightbox to a computer workstation, teleradiology is no longer an extraordinary task; it is actually one of the normal instruments of conducting radiological work. The parallel development of digital imaging and PACS has thus given a push forward for teleradiology, too (Benjamin et al. 2010). 20 2.3 Required components of a teleradiology system A basic teleradiology system consists of three major components: 1. 2. 3. An image-sending station A transmission network A receiving/image review station. Patient images are electronically encoded in a digital format at the sending station, sent via the transmission network, and received, viewed and possibly stored at the review station. (Kumar S., 2008.) A generic teleradiology set-up is shown in figure 1. Satellite Microwaves Film Digitizer, (Video, Laser, CCD) Mobile data Images Internet DICOM Images, TXT Teleradiology workstation or Direct Digital Capture Formated image data / transmission system HIS/ RIS/ EPR Patientrelated information WAN Network access controller Security Controller DSL ISDN Network access controller Management Software Security Controller Storage / PACS Phone Report transmitter Report receiver Fax/Phone SENDING SITE Phone line DATA NETWORK Links: billing, HIS, RIS Fax/Phone RECEIVING SITE Fig. 1. Generic Teleradiology component set-up (Huang 2010, Kalyanpur et al. 2004, Kumar 2008, Lee et al. 1999). The components are explained in more detail in the adjacent text. 2.3.1 Image-sending station At the image-sending station images are captured into digital format with a digitizer, if originally stored on x-ray film. A laser film digitizer or a CCD (charge coupled device) film scanner is usually used for this purpose. Today, with most of the images already in digital DICOM (digital imaging and communication in 21 medicine) format, a direct digital output from digital modalities (CR, CT, DF, DM, DR, DSA, MR, US) is utilized. (Huang 2010.) For older devices, also video frame grabbers are used (Kroeker et al. 2000). If a digital connection to HIS and RIS exists, this information is also combined with relevant images and sent to the receiving site (Huang 2010). Otherwise a fax connection or even a telephone connection can be used to communicate the relevant patient information (Dunne & Dunne 2004). For image and related data communication, it is advantageous to convert the data to the industry standard because multiple vendors’ equipment is used in the teleradiology chain. Specifically, the DICOM and TCP/IP (Transmission control protocol/Internet protocol) standards should be used for image data communication. (Huang 2010.) 2.3.2 Transmission network Communication networks are essential for teleradiology. The networks can basically be wired or wireless. The simplest wired solutions are telephone lines or digital ISDN (integrated services digital network) lines that can be used as pointto-point connections with the help of a modem. They are also known as dial-up connections, because they are established as need arises. Their security requirements are less stringent, because there are no other partners in the communication pathway. Their usable speed starts from 9.6 kbit/s for a modem connection and 64 kbit/s for a one-channel ISDN connection. ISDN connections can be multiplied; a typical setting uses 2 to 6 parallel channels with a speed of 128 to 384 kbits/s. (Lear et al. 1989, Stewart et al. 1992, Huang 2010.) Wide area network (WAN) connections are fixed connections between local area networks (LAN) in different institutions. They typically provide higher speeds and more reliable connections. A lower-cost solution to WAN connection is using DSL (Digital Subscriber Line) at the customer site of the connection. A specific type of WAN connection is the Internet, which technically in most cases uses WAN backbone, but includes the possibility to establish connections more easily, even on an ad hoc basis. (Arenson et al. 2000, Huang 2010.) Wireless network connections for teleradiology include satellite transmission, microwave transmission and mobile phones connections. Satellite connections are used directly or as part of WAN connections. Microwave transmission usually involves direct point-to-point connections. Mobile phone network connections are 22 based on the data transfer capability of the operator lines and are usually connected to LAN at the sending site. (Lamminen 1999, Huang 2010.) The recent developments of WAN and mobile phone networks are discussed in more detail in the next chapter. 2.3.3 Image review station For a teleradiology review system, workstations with high-end 1024 × 1024 pixel resolution (1 K) monitors for cross-sectional images and workstations with multiple 2048 × 2048 pixel resolution (2 K) monitors have been recommended for CR and DR (digital radiography) primary diagnosis (Huang 2010). As the radiology workflow has turned digital, the American Association of Physicists in Medicine (AAPM) has published requirements for image reading displays in general PACS environment, and the current recommendations include not only definitions of display resolution, but also definitions of other display characteristics, such as luminance, as well as recommendations for image reading environments and quality assurance including DICOM calibration of the displays (Samei et al. 2005a, Samei et al. 2005b). The AAPM recommendations make no difference between teleradiology or local reading, but they do distinguish quality requirements between displays used for primary diagnosis and secondary consultation. In Finland, the Radiation and Nuclear Safety Authority (STUK) has published a Guide for Quality Control of X-ray equipment in Health Care (2008) which covers also image display monitors and required steps in quality maintenance. More recently, Liukkonen has published a thesis (2010) which can be used as a primer for display requirements. The workstation should include local storage for images and user-friendly software to manage and display the images and relevant patient data if available. Ideally, the teleradiology workstation should be connected to the local PACS for image store and comparison image retrieval. For the expert centre, administrative tools for reporting and billing are also necessary. (Huang 2010.) 2.3.4 Development of workstation and digitizer technology In the early days of digital medical images, the concept of high-quality teleradiology consisted of laser film scanners with large scanning matrix like 2048 × 2048 pixel resolution and high-end computer workstations together with fast connection lines. (Batnitzky et al. 1990, Stewart et al. 1992, Goldberg et al. 23 1993.) The costs of such a system could be considerable up to the level of USD 200,000. (Dwyer et al. 1991). The development of computer and software technology made it possible to achieve acceptable computational power with more commonly available personal computers (PC) and to substitute dedicated software with more general components (Yamamoto et al. 1993, O'Sullivan et al. 1997). CCD film scanner technology has also progressed so as to produce image quality that is getting close to the quality of the more expensive laser film scanners (Boland 1998, Calder et al. 1999). 2.4 Recent communication technology developments Teleradiology requires efficient and affordable communication networks. The access to fast computer networks is now easier than before, and interestingly enough, open networks like the Internet as well as mobile data communication provide new possibilities. 2.4.1 Wired connections Teleradiology connections first started with modem-based telephone lines, after which they developed into digital ISDN connections and the high-capacity digital networks that are readily available today (Kumar 2008). Institutions can hire fast WAN connections speeds in the magnitude of 1 to 155 Mbit/s and more between their local networks. The costs rise in proportion to network speed. These WAN connections are established together with a telecom operator, and appropriate security measures are taken. There is a possibility to purchase a guaranteed bandwidth or quality of service (QoS) in order to have a constant transfer speed. (Huang 2010.) These high-capacity networks are mostly used for image traffic between PACS archives and also for regional teleradiology and image transfer networks (Bradley 2008). Private persons can also lease DSL (Digital Subscriber Line) or optic cable connections to their homes with capacities from 256 kbit/s to 20 Mbit/s. Even though the costs are lower, these lines are still suitable for teleradiology. (Huang 2010.) 24 2.4.2 Connection of networks; the Internet The Internet is a global system of interconnected computer networks that use the standard Internet Protocol Suite (TCP/IP) to serve billions of users worldwide. It is a network of networks that consists of millions of private, public, academic, business and government networks of local to global scope that are linked by a broad array of electronic and optical networking technologies. The Internet carries a vast array of information resources and services, most notably the inter-linked hypertext documents of the World Wide Web (WWW) originally developed in 1989 at CERN, the European Organization for Nuclear Research, and the infrastructure to support electronic mail. (Ahonen 2008, Bergh et al. 2000.) Started as the US military network Arpanet in the late 1960s, the Internet with its protocols has now achieved a global position. In Finland, university networks were started in the 1970s, and in 1983 the Finnish University and Research Network (Funet) project was founded. The Nordic countries were forerunners in Europe by establishing a joint NORDUnet research network in 1985. The NORDUnet network had the benefit of being neutral to networking protocols, which enabled researchers to use the Internet ahead of central European countries like Germany, England, France or the Netherlands. In 1988, NORDunet was actually the first network outside the USA to have a connection to Internet. Researches in the Nordic countries had thus better opportunities for international collaboration. (Ahonen 2008.) After the Internet was made publicly available outside the research community as well in the mid-1990s, there was increasing interest in its use for medical information transfer. For teleradiology, the possible benefits of the Internet are related to the ease of making connections, even internationally (Boland 1998). There is however no QoS on Internet connections (Huang 2010). Because of slow speeds and security concerns, the Internet and more specifically, WWW, was seen a tool for educational material delivery and discussion (Scarsbrook et al. 2005). Patient treatment over the network was considered to require more advanced secured tunnelling protocols and more efficient ways of forwarding large amounts of data (Bergh et al. 2003). 25 2.4.3 Mobile communication Mobile data networks Wireless networks became meaningful for medical purposes after they covered a reasonably large area. The Nordic Mobile Telephone (NMT) network was the first fully automatic wireless network that provided multinational roaming from country to country and could be used for data transfer (Ali-Yrkkö & Hermans 2004, Reponen & Niinimäki 2006). One of the benefits of the NMT network was that its 450 Mhz variant also covered more distant areas of Finland (Judén & Kuokkanen 2000). Although the GSM (global system for mobile communications) developed in Europe is the most widely distributed second-generation mobile network, different technologies are used in the US and Japan, making them incompatible (Istepanian et al. 1998, Istepanian et al. 1999). Interestingly, the first GSM call ever was made in Finland in 1991 (Elisa Oyj 2004, GSM Association 2010). One benefit of GSM for medical use is that digital data transfer was defined as one of its basic functions from the beginning. The main limitation of circuit switched GSM data networks is the data transfer speed, which is realized in commercial services in the magnitude of 9.6 to 14.4 kbit/s (high-speed), although data rates up to 57.6 kbit/s would be theoretically possible (Joutjärvi et al. 2000, Tachakra 2003). The next generation of data networks, the so-called 2.5 generation, namely packet-switched GPRS (general packet radio service) and its variant EDGE (enhanced data rates for GSM evolution), provide theoretical maximum data transmission speeds of up to 114 and 473.6 kbit/s, respectively. In practice, the effective speed is lower. The increase in speed is substantial compared to original GSM data transfer, enabling much faster data transmission. Another benefit of these packet-switched networks is that they allow the usage of Internet protocols in data transmission, which makes system design much easier for mobile data applications. Finally, third-generation (3G) UMTS (universal mobile telecommunications system) networks have developed data transfer speeds to a level that makes mobile Internet usage feasible. (Ericsson 2009.) All these variants of data networks are currently available, GPRS coverage being available practically in all parts of Finland and in most European countries, its faster EDGE variant having a slightly more limited coverage. Fast 3G 26 networks, even though available world-wide, are still concentrated in the most densely populated regions. (GSM Association 2009.) Handheld terminal devices Thanks to advances in processor and display technology, handheld computers or personal digital assistants (PDAs) gained a market in the 1990s (Yamamoto 2000, Pyrros & Yaghmai 2008). Mostly equipped with a black and white display and a slow processor, they were capable of displaying elementary images. The variants included devices from manufactures like Psion, Ericsson, Palm and Sharp. Networking capacity was added to some of the devices via a cable modem or a network card, or with a GSM data card or a cable connection to a mobile phone (Tachakra 2003). The concept thus resembled that of a laptop computer with networking capabilities. A revolution took place in 1996 when Nokia Corporation introduced the first integrated communicator device, which had a handheld PDA-type computer and a GSM phone with digital data transfer capabilities in one package (PDAdb.net 2007). The phone functions included advanced SMS messaging, and the PDA functions included office solutions as well as typical PDA applications like calendar and voice recording. The most important factor was the possibility to make new software applications for the device and also to guide it remotely. This opened the possibility of using it as a terminal device for various computer and data network applications. The display was sharp and larger than in ordinary mobile phones. The limitation was its grey scale, but with proper software to compensate for this, it was possible to use it for displaying grey-scale images. 2.5 DICOM-standard Radiology has the benefit of standing in the crossroads of medicine and technology. The cost of equipment is high, so the investments should be economically feasible. In the digital imaging domain, it was very soon noticed that no single vendor could provide a comprehensive imaging system for customers. The integration of different imaging systems is expensive and timeconsuming and the business prognosis was poor if customers could not trust the vendors. This prompted the ACR and the National Electrical Manufacturers Association (NEMA), with input from various vendors, academia, and industry groups, to work together to come up with standards of interoperability in medical 27 imaging from 1983 on, and published finally the DICOM standard in 1993 (NEMA 2010). DICOM is a standard that is a framework for medical-imaging communication. Based on the open system interconnect (OSI) reference model, which defines a 7-layer protocol, DICOM is an application-level standard, meaning that it exists inside layer 7 (the uppermost layer). It is referred to as “version 3.0” because it replaces versions 1.0 (published in 1985) and 2.0 of the standard (published in 1988) previously issued by ACR and NEMA known as the “ACR-NEMA” standard (Kiuru 1993, NEMA 2010). DICOM provides standardized formats for images, a common information model, application service definitions and protocols for communication in medical imaging in general, not only in radiology (NEMA 2009). DICOM is a cornerstone standard in medical imaging, and very rapidly after its launch in 1993 it replaced local European and Japanese standardization initiatives. It is now developed in liaison with other standardization organizations including CEN TC251 (European Committee for Standardization, Technical Committee 251) in Europe and JIRA (Japan Industries Association of Radiological Systems) in Japan, with review input from other organizations including the IEEE (Institute of Electrical and Electronics Engineers), HL7 (Health level 7) and ANSI (American National Standards Institute) in the USA. (NEMA 2009.) One of the factors leading to rapid implementation of the DICOM standard was commitment on the part of end-users. When DICOM became a requirement in released calls for tender, new equipment and software soon complied with the standard and made more complicated integration possible after the latter half of the 1990s. (Kotter & Langer 1998, Kinsey et al. 2000.) Because the standard enabled various vendors to work together, the medical imaging sector has been a forerunner in digital patient data usage. A common standard has increased the market sector and has made it possible for so-called third party vendors to enter the market with new innovations. 2.5.1 IHE integration profiles DICOM is in no way perfect; there is quite a lot of freedom in its implementation, which is why a new initiative called IHE (Integrating the Healthcare Enterprise) has been established (Oosterwijk 2004). The purpose of IHE is not to create a new standard, but to develop further so-called integration profiles, which describe 28 how everyday tasks are best performed with existing equipment and available standards, such as the imaging standard DICOM and the medical record messaging standard HL7 (Channin 2001). IHE has reached a critical mass in product support, enabling a significantly higher level of interoperability compared to former system installations. Integration profiles in particular allow broad implementation and useful extension of existing and newly developed IT systems in healthcare (Wein et al. 2005). IHE is organized across a growing number of clinical and operational domains. Each domain produces its own set of Technical Framework documents, in close coordination with other IHE domains. Currently in 2010 there are nine different domains. In the radiology domain there are 19 IHE profiles that have been published in trial implementation or final text versions. For example a profile called Cross-enterprise Document Sharing for Imaging (XDS-I) extends XDS-profile (which registers and shares electronic health record documents between healthcare enterprises) to share images, diagnostic reports and related information across a group of care sites (IHE 2010). Titles of the radiology domain profiles are presented in Table 1. Table 1. Titles of the published radiology domain IHE integration and interoperability profiles in 2010. Detailed up-to-date description of the functions of each profile can be found at http://www.ihe.net/Profiles/. Content Profiles Workflow and Infrastructure Profiles Consistent Presentation of Images (CPI) Scheduled Workflow (SWF) Key Image Note (KIN) Patient Information Reconciliation (PIR) Nuclear Medicine Image (NM) Post-Processing Workflow (PWF) Mammography Image (MAMMO) Reporting Workflow (RWF) Evidence Documents (ED) Cross-enterprise Document Sharing for Imaging (XDS-I) Simple Image and Numeric Report (SINR) Portable Data for Imaging (PDI) Teaching File and Clinical Trial Export (TCE) Import Reconciliation Workflow (IRWF) Image Fusion (FUS) Presentation of Grouped Procedures (PGP) Charge Posting (CHG) Access to Radiology Information (ARI) Audit Trail and Node Authentication (ATNA) Product vendors test the interoperability of their products in so called Connectathon events. The results are then published in an open database (IHE Europe 2010). 29 Development of these standards and integration profiles has been important in also making teleradiology an extension of existing imaging networks, thus helping its distribution as a health care service (Hussein et al. 2004). 2.6 Medical Devices Directive Another form of coordinating regulation, which will in the future have an impact on software used for teleradiology is the Medical Devices Directive (MDD). It is a safety regulation that discusses the responsibilities of the manufacturer, the user and the authorities. More recently, Directive 2007/47/EC (2007) amended the definition of the term “medical device” used in Directives 90/385/EEC (1990) and 93/42/EEC (1993). Recital 6 of Directive 2007/47/EC (2007) states that “it is necessary to clarify that software in its own right, when specifically intended by the manufacturer to be used for one or more of the medical purposes set out in the definition of a medical device, is a medical device”. In Finland, the corresponding update in legislation of medical devices came into effect on 1st July 2010 (L 629/2010). An important implication of these above mentioned regulations is that software products in medical device categories should fulfil the essential requirements and their risk class has to be classified. Depending on risk class, there are different evaluation steps before the product can receive a CE mark (European conformance marking) and enter the market. The regulations promote the safety and quality of the products and give users a mechanism to report any serious problems. The Swedish Medical Products Agency has published a guidance report (Läkemedelsverket 2009) on how to classify various IT systems. This report has been a basis for the work of the European Union (EU) Medical Devices Expert Group, which is preparing European guidelines on this topic. According to the guidance report, telemedicine programs, including teleradiology programs, are considered to be not less than risk class I medical devices. The same implies to software products running PACS when images are viewed and transmitted and to RIS when information is used for treatment planning of individual patients. 2.7 Image compression Digital images can be compressed in order to decrease the file size. This has the benefit of decreasing either the storage space or the bandwidth needed. The latter 30 aspect is related to required network capacity and image transfer times. The general rule is that the more capacity is needed, the more it costs and the more time image transfer takes. (Logeswaran 2008.) As access to affordable bandwidth has increased, the importance of image compression has decreased in normal consultation settings. In most the cases, a full uncompressed DICOM radiographic image can be sent for a consultation. However, image compression is still an issue if wireless networks are used for mobile terminals and if images are sent over otherwise narrow bandwidths, such as satellite connections. In addition, new modalities like multislice CT, MRI and DM produce image packages that are transmitted slowly if uncompressed. (Logeswaran 2008.) There are basically two types of compression, lossless and lossy (Kajiwara 1992). The first typically results in 1:2 or 1:3 gain in image size, while lossy compression for medical purposes can be used without noticeable visual loss with resolutions such as 1:7 to 1:40, depending on the modality and type of images (Savcenko et al. 1998, Savcenko et al. 2000). For lossy compression, the most widely used are JPEG (Joint Photographers Expert Group) compression and wavelett compression. Fractal compression is not commonly used for medical purposes because of its inferior performance. (Ricke et al. 1998, Kotter et al. 2003.) The artefacts generated with compression types vary, and according to reports, influence their suitability for different types of clinical images. Even though wavelet compression usually enables a higher compression rate compared to JPEG compression, this is not always true in clinical settings. (Sung et al. 2002, Schulze et al. 2008.) Many studies have shown that properly chosen lossy compression has no effect on diagnostic quality (Lo & Huang 1986, Ishigaki et al. 1990, MacMahon et al. 1991). In practice, PACS vendors have implemented compression in their clinical viewer software in order to decrease bandwidth requirements for in-house networks (Logeswaran 2008). New recommendations on the use of image compression have been published by three radiological societies over the last two years 2008–2009: – – Royal College of Radiologists (RCR, UK) “The adoption of lossy data compression for the purpose of clinical interpretation” (RCR 2008) German Röntgen Society (DRG, Germany) “Compression of Digital Images in Radiology - Results of a Consensus Conference” (Loose et al. 2009) 31 – Canadian Association of Radiologists (CAR, Canada) “Pan-Canadian Evaluation of Irreversible Compression Ratios (“Lossy” Compression) for Development of National Guidelines” (Koff et al. 2009) These recommendations have some similarities, e.g. the use of image compression is possible without losing relevant clinical informations. There are slight differences in accepted compression ratios. Based on these papers, users should be able to implement workflows using “lossy” image compression. However, that there are still some concerns due to different aspects, e.g. responsibility, algorithms, post processing. Therefore ESR has initiated a process to discuss these open issues including well recognized international experts in this field. The estimated outcome is an official ESR recommendation (ESR 2010). This activity has been coordinated by the ESR subcommittee for “Information and Communication”. 2.8 Data safety Patient information should be kept secret and unmodified, but the data should be available when needed in patient care. The privacy policy means that only authorized persons are allowed to see personal health data. (Ruotsalainen 2010.) These requirements are common in any type of teleradiology. Telemedicine widens the access to sensitive data and therefore data security development is an essential part of telemedicine projects. Publicly available security evaluation methods provide a good starting point for security work. A systematic approach to data security development facilitates the process, but does not give answers to all practical questions. (Niinimäki et al. 2001.) The security and privacy protection of any teleradiology system must be planned in advance, and the necessary security and privacy enhancing tools should be selected based on risk analysis and requirements set down by legislation (Ruotsalainen 2010). There is, however, a trade-off between security measures and system performance (Engelmann 1999, Zhang 2007). Wireless security specifically has three goals: (1) restricting wireless access to authorized users, (2) controlling which devices are allowed access to the host network, and (3) encrypting wireless data transmission to prevent eavesdropping, modification or insertion of data with respect to the wireless data stream. Server and client security (including PCs, laptops, PDAs etc) is a separate, but equally important, security objective. (Pyrros & Yaghmai 2008). 32 2.9 Work process As mentioned earlier, the first clinical teleradiology experiments were based on television technology, which mostly required simultaneous presence of the participants. A major logistic step forward was the store-and-forward approach, which simplified operation by eliminating the need for all parties – patients, providers and other support staff – to be present simultaneously at both sites (Thrall 2007a). In terms of systems design, teleradiology can be seen as one application area of interactive or computer-supported cooperative work (CSCW) systems. Within systems design, it is quite common to view such systems in terms of workflow, often accompanied by a chain of tasks view of work (Abbot & Sarin 1994). These identifiable, usually system use-related tasks make up only part of the teleradiology service and the time spent on making it work. It is not sufficient to study only these isolated tasks, however; an investigation of actual, everyday routine activities is needed in order to understand the contextual details of work and interaction in ways that support the usability of the system. Even though technical (Uldal et al. 1997, Glinkowski & Kornacki 2002, Struber & Tichon 2003), clinical (Goh et al. 1997a, Goh et al. 1997b, Kiuru et al. 2002, Stranzinger et al. 2003, Daucourt et al. 2005) and economic (Bergmo 1996, Halvorsen & Kristiansen 1996, Johansen & Breivik 2004, Daucourt et al. 2006, Kreutzer et al. 2008, Plathow et al. 2005) aspects of teleradiology have been investigated intensively, there are not many studies with a work practice-oriented focus describing real-life teleradiology work between primary and secondary care (Karasti 1997, Karasti 2001). 2.10 Motivation of teleradiology 2.10.1 Effectiveness of teleradiology Teleradiology has been seen as being economically feasible in Norway in a situation where it replaced a visiting radiology service to a remote hospital, the examinations consisting of plain radiographs (8,000 examinations per year) (Bergmo 1996). Another early example from Austria showed that teleradiology of emergency CT examinations from a regional hospital to a university hospital was economically feasible compared to patient transportation even with a reasonable 33 low number of examinations (121 examinations over a period of 13 months) (Stoeger et al. 1997). Similar benefits have been reported by Goh et al. (Goh et al. 1997a) on the role of teleradiology in the management of inter-hospital transfer of neurosurgical patients. They found that for patients referred with teleradiologic images, unnecessary transfers were reduced by 21%. In a large prospective series from Aquitaine France, all teleradiology transmissions in 15 hospitals and over a period of one year were examined. There were 664 (90%) transmissions that met the inclusion criteria. Of these, 562 (85%) were for emergency care and 102 (15%) for non-emergency care. In emergency care, the pathologies most often associated with a remote consultation were cerebral pathologies (88%) and traumatic spinal pathologies (8%); the proportion of avoided transfers was 48%. In non-emergency care, the specialties most often concerned with remote consultations were neurology/neurosurgery (36%), cardiology and pulmonary diseases (17%) and gastroenterology (14%). Transfer was avoided for 37% and hospitalization for 12% of the patients. An additional consultation occurred after remote consultation for 2% of the patients (Daucourt et al. 2005). A cost-minimization study enabled a comparison of care procedures following the use of the network with those that would have been implemented without the network. Annual savings can be estimated at EUR 102,779 for the Aquitaine area (Daucourt et al. 2006). The results confirm the effectiveness of an inter-hospital teleradiology network. (Daucourt et al. 2005.) In a primary care setting, a teleradiology service linking a general practice in rural Norway with the local hospital in the nearest town was evaluated, and savings on travelling expenditure were registered for all patients who underwent elective examinations during the first year of service. Over a four-month period in 2002, it was recorded whether patients undergoing emergency examinations were taken in at the local general practice or referred to the hospital. The method employed was a cost-minimization analysis in which the costs of teleradiology are compared to the costs incurred when patients go to hospital for a radiological examination. On the basis of data for 3,006 patients, the analysis showed that on an annual basis the service resulted in savings of NOK 1 million (USD 160,000). According to the authors, teleradiology is most likely to be cost-effective when annual patient load and travel costs are high, together with relatively low investment costs. (Johansen & Breivik 2004.) In Finland, the clinical effectiveness of primary care teleradiology has been well demonstrated (Kiuru et al. 2002). In the study, 685 plain film examinations 34 were sent from a primary care centre to a university hospital via an ISDN connection for a radiological report. The study was conducted in two phases: during phase 1 (446 cases), general practitioners (GPs) selected the examinations, and during phase 2 (239 cases), all consecutive examinations were transmitted. In at least one-third of all cases, teleradiology helped with the diagnosis, although completely new diagnoses were less common. An effect on treatment was noted in 15% and on prognosis in 5% of the cases. However, an appropriate consultation level is required for these positive effects. Adequate accuracy and patient safety cannot be achieved if the examinations sent for radiological reporting are preselected by a GP. In their systematic review of telemedicine articles Roine at al. were able to show that one the strongest pieces of evidence of the effectiveness of telemedicine is within teleradiology, especially teleneuroradiology (Roine et al. 2001). 2.10.2 Quality aspects A proper quality assessment should be included in teleradiology service in order to assure users that remote interpretation is as accurate and fluent as local image reading (Wong et al. 2005, Branstetter 2007). A Tyrolean ‘four-column model of quality management in telemedicine’ was introduced to ensure a global view of the project and to avoid mistakes. In teleradiology, a 12-step workflow was developed, describing the medical responsibilities at each stage. The investigators found that the defined teleradiology workflow and the technical equipment for data security and data exchange worked without problems in over 79% of a total of 424 cases. To ensure continuous quality assurance, the whole teleradiology workflow was ISO 9001:2000 certified. (Soegner et al. 2003b.) In another study, careful analysis of errors resulted in reduction of the number of mistakes (mostly minor workflow errors), from 23 errors per month in January 2002 to 9 errors per month in January 2003. The authors concluded that continuous cross-checking of the workflow and training of the employees involved guaranteed a better standard of teleradiology. (Soegner et al. 2003a.) 35 2.11 Use of teleradiology Several factors – including the prevailing shortage of radiologists, the increasing use of advanced imaging methods, the consolidation of hospitals into regional delivery systems, and heightened expectations of patients and referring physicians for timely service – have fostered the increasing use of teleradiology. These factors have also helped underwrite the creation of new and potentially disruptive business models for service delivery that can be viewed as threats, opportunities, or both, but which cannot be ignored. (Thrall 2007a.) 2.11.1 Point to point consultations Teleradiology has traditionally been used for secondary opinion or for off-hours emergency consultations (DeCorato et al. 1995, Wilson 1996). It has also been proposed for educational purposes (Thrall 2007a, Canadian Association of Radiologists 2008). A rather new aspect in teleradiology is outsourcing the workload, where a major part of the service is given away from regular staff to an external expertise centre. In the US, a new term, “nighthawking” has been coined, because off-hour consultations by external experts are such a common practice. (Boland 2008.) At the 2007 meeting of the Radiological Society of North America there were 41 teleradiology companies that provided nighttime coverage, daytime coverage, or all of the above (Bradley 2008). According to another study, 44% of all radiology practices in the United States reported using external off-hours teleradiology services in 2007. These practices included 45% of all U.S. radiologists. Out-of-practice teleradiology had been used by 15% of practices in 2003, so the amount has tripled. (Lewis 2009.) In the EU area there is service provider based in Spain that provides local as well as international services (Eklof et al. 2007). 2.11.2 Regional service concept Regional access to a joint PACS archive actually means teleradiology workflow, if the images are interpreted in a different location from the imaging site. This concept makes it possible to build up a virtual hospital, where even small sites can enjoy sub-specialist consultations or where sub-specialist expertise can be used more efficiently without physical boundaries. Primary health care centres can have access to radiology consultation as if they had this service on their own 36 premises. An additional benefit is access to older images and thus avoidance of excess radiation exposure. (Harno et al. 2002, Harno 2004, Vogl 2005.) For image interpretation it is also important to have comparison images from other sites available; this allows physicians to follow up pathologic changes over time, which results in more accurate diagnosis. One major regional PACS installation in Finland is HUSpacs, consisting of the images of 21 hospitals and related primary health care centres (Pohjonen et al. 2004). In Austria, the Insbruck region has built a comprehensive telemedicine installation connecting the university clinic with a district hospital (Soegner et al. 2003b). Both of these examples implement remote interpretation of images. Integrating primary health care centres to secondary care via information systems can create a virtual electronic health record and change service models (Harno et al. 2002). When associated with viewing of patient data through the link directory, interactive eConsultations enable supervised care leading to the reduction of outpatient visits and more timely appointments. In the HUSpacs information system the link directory is an interface containing links pointing to the actual patient data located in remote information systems. The original data including images can be viewed with a web browser, but data can be accessed only with the patient's informed consent. This means that the citizen has an active role in deciding on the use of his medical information (Harno 2004). In this manner the link directory system has implemented the European data protection directive (Directive 95/46/EC 1995). An interesting idea has been proposed by Wootton, stating that one advantage of teleradiology would be omitting the need to maintain specialist staff in hospitals where the volume of radiology may not justify it (Wootton 2001). 2.11.3 Mobile teleradiology Wireless transfer of radiology images to a portable computer was reported early in emergency medicine using analogue cellphone technology (Yamamoto 1995). With the advance of digital cellular phones and worldwide digital networks image transmission has been extended to major catastrophe sites and for purposes such as delivering information needed in post mortem recognition of bodies (Salo et al. 2007). Wireless technology is identical to conventional teleradiology, with the obvious difference of using a radiofrequency wireless network for the digital communication of radiographic images. There are four major disadvantages with 37 wireless technologies: reliability, cost, security and speed. The obvious major advantage of such technology is the ability to view images virtually anywhere, from a hospital patient room to the quagmire of the battlefield. In addition, the costs of wireless networks and their speed have become insignificant issues: many would in fact argue that deploying a wireless network saves costs, as it does not require the expense of cabling. (Pyrros & Yaghmai 2008, Thrall 2007a, Thrall 2007b.) Wireless devices have been utilized with wireless broadband technology within the hospital environment in ward rounds. In a Finnish study a mobile Tablet-PC/Webpad- based system with WLAN (Wireless Local Area Network) connectivity was used to access both textual and image data in the wards. The data access was fast and comparable with terminals connected to the ordinary hospital network. The main limitation of WLAN was that outside the hospital campus area it was only available at so-called service points, “hot spots”. WLANbased image review was thus not considered suitable for health professionals who are on duty and have to be able to connect to the hospital at any time. (Pagani et al. 2003.) A new type of equipment class entered the market when pocket computers emerged, and to demonstrate the feasibility of wireless pocket teleradiology, brain CT scan images of five neurosurgical emergency cases were received on a pocket computer via a wireless modem link (Yamamoto & Williams 2000). 2.11.4 International teleradiology consultations International teleradiology was first developed in a military setting. The first published article describes how an US army team lead by Lt. Col. Fred Goeringer, MSC, deployed a CT scanner to a transportable hospital in the Saudi Arabian desert during the first Gulf War in 1991. The images were transmitted via satellite and telephone links to Brooke Army Medical Center in San Antonio, Texas, for specialist radiology consultations. Another objective was to validate the concept of distant interpretation of images obtained on the battlefield. According to the report, the scan quality was excellent and the scans proved clinically important in patient management. (Cawthon et al. 1991.) This pilot showed great promise for both combat and civilian teleradiology and led to further development (Mun et al. 1998). The Internet opened up new possibilities for international consultation possibilities. The Nordic countries were among the forerunners in utilizing 38 Internet connections (Ahonen 2008). The new route was at that time tested for medical purposes. Data safety was a concern at the beginning, but patient data anonymization, secured network connections like VPN (virtual private network) and SLL (secure socket layer) technology or data encryption were gradually adopted in public network connections and international connections as well. (Wunderbaldinger et al. 1999, Seibel et al. 2000, Niinimäki et al. 2001, Alaoui et al. 2003, Bergh et al. 2004.) Today, technical aspects do not necessarily restrict international image transfer. Images are stored in DICOM format and can thus be transferred with DICOM command to capable destinations. What is more complicated is the transfer of equally important RIS information and EPR information. This is related both to lack of commonly used standards in this area as well as possible language issues. (Arenson et al. 2000, Siegel & Reiner 2002, ESR 2004.) 2.12 Concept of eHealth Utilization of ICT (information and communication technology) in health care has moved from individual projects and services to a more comprehensive model. Instead of telemedicine, terms like telehealth, eHealth, on-line health and most recently connected health (Microsoft Corporation 2009) have emerged. eHealth is an umbrella term that gathers together ICT usage in health care (World Health Organization 2010). It includes major infrastructure services and at the other end also services targeted to individual citizens. According to the EU eEurope action plan “An information society for all” eHealth could also seen as a derivative of eServices, in line with similar terms like eCommerce or eGovernance (COM/2002/0263 final). The main political goal is to deliver more accessible and better services to citizens. eHealth has been strongly emphasized by the European Commission since 2003, when the first eHealth ministerial meeting was held and the first Ministerial Declaration (2003) on the topic was given. Since then, the European Commission has encouraged member states to include eHealth in their national health policies. The WHO has also created its own world wide eHealth strategy (World Health Organization 2010). 39 2.12.1 National initiatives for electronic patient documents Finland is a good example of a country that has been an early adopter of eHealth policies. The information management infrastructure of Finnish health care on a national level will be based on an internationally unique centralized electronic archive. The operative use of EPRs will take place on a regional or local level. (Hämäläinen et al. 2009, Winblad et al. 2010.) The core of the national Finnish health ICT infrastructure will reside in the national digital archive for patient documents. It is since 2007 based on legislation (L 159/2007) and its implementation will be mandatory for all health care providers in the public sector. For private providers of physician services the implementation will be mandatory for those using electronic documentation of patient data. The national architecture consists of local EPRs using a common data structure and technical standards, the national eArchive in which all EPRs are made available online subject to patients’ consent (L 159/2007 2007). In addition, the archive will include a national ePrescription database based on act on electronic prescription (L 61/2007 2007), and an eView for citizens, enabling them to access their own patient data and log data (L 159/2007 2007). There are several challenges to be overcome with the national archiving system. It is a technologically complex system that has to be implemented alongside the everyday service routines of health care. In order to work on a nationwide level, interoperability of systems is crucial. In addition, the volumes of information transfer, for instance in radiology, may be very large and overload the data networks. The system requires that health care employees use an electronic signature and learn new ways to work, such as documenting patient information in a structured manner. (Winblad et al. 2010.) 2.12.2 EU cross border telemedicine initiatives Since the beginning of 2000s, the European Commission has become interested in the possibilities of cross-border telemedicine. The topic has been discussed in the EU-funded projects Baltic eHealth and R-Bay (Ross et al. 2010). In the Baltic eHealth project teleradiology was utilized between the Baltic states and Denmark. The image interpretation results were dictated in the native languages of the radiologists and translated into Danish with a computer-based structured form. Legal issues concerning cross-border telemedicine were also investigated in the 40 project. In the R-Bay project a more general concept of an international teleradiology marketplace was studied. More recently, the EU has initiated the epSOS project (Smart Open Services for European Patients), where interoperability issues are investigated between the electronic patient record systems of various European countries in a more general setting of cross-border telemedicine (epSOS 2008). The EU Commission has promoted cross-border telemedicine and eHealth in ministerial meetings, given out various statements: a Commission Recommendation on cross-border interoperability of electronic health record systems (C(2008)3282), a Commission Communication on telemedicine for the benefit of patients, healthcare systems and society (COM/2008/0689 final) and a Ministerial declaration (2010) on European co-operation on eHealth. The EU parliament has been also discussing incorporating definitions of cross-border telemedicine into the directives on the mobility of services (Commission Proposal for a directive 2008/0142) which will be voted in January 2011. The European Society of Radiology (ESR) has published two guideline “white papers” concerning teleradiology (ESR 2004, ESR 2006). In these papers, teleradiology is seen as a feasible opportunity, but ESR stresses the quality aspects. In relation to directives discussing cross-border telemedicine, ESR reminds that teleradiology is a medical act and should be treated accordingly. 2.12.3 Teleradiology as a part of eHealth services In the future health care environment teleradiology has a strong role as one of the services given (Shieh et al. 2008). It is a means of delivering care, but not a medical discipline in itself. One key issue is the well-established integration to other health information systems, especially RIS (Roberson & Shieh 1998). From the point of view of users, usability and reliable functionality are key issues (Shieh & Roberson 1999). Health care services are becoming more ubiquitous (Kyriacou et al. 2009). Experts need to be reachable anytime and anywhere, and they require the same amount of information they rely on while working within hospitals. For teleradiology, the technological possibilities with more capable handheld devices and even faster wireless networks are no longer an obstacle. Mobile health care systems (mHealth) have been said to be the next major step forward in telemedicine. (Voskarides et al. 2002, Herscovici et al. 2007, Waegemann 2010.) 41 Virtual hospital is already a existing possibility (Jones 2000, Graschew et al. 2006). Various institutions can be linked together and workflows can be synchronized (Bradley 2008). This requires hard work, however, as integration standards are not well developed or not adopted to a satisfactory degree in other HIS compared to medical imaging systems. For teleradiology purposes, the existing infrastructure is for the most part sufficient if RIS and HIS information can be transmitted. A common regional information system could help solve many of the problems that exist today. (Harno et al. 2002, Harno 2004.) Outsourcing of work is a solution to distribute the burden of excess workload and perhaps a tool to connect sub-specialist radiologists with their own patients (Bradley 2008). A new structure for the radiology department is also needed. The risks are also obvious – how to cope with quality issues and keep up interest also in the boring, bulk-type interpretation work? 42 3 Purpose of the study The purpose of this study was to evaluate the use and implementation of teleradiology, namely: 1. 2. 3. 4. 5. 6. whether a low-end technology teleradiology system could combine acceptable image quality with reasonable costs and serve primary health centres and local hospitals for image consultation whether there will be noticeable changes in work processes and what are the most important work practice issues when two different institutions start to work together, whether international teleconsultations using data networks could link together different institutions in a feasible manner, whether new ubiquitous communication tools, such as mobile data communication, are suitable for radiological purposes, whether smartphones could be used for image interpretations, and finally, to chart the usage status of teleradiology and digital radiology in the electronic health care (eHealth) domain in Finland in general. 43 44 4 Subjects and methods 4.1 Low-end technology teleradiology The hardware consisted of a standard PC computer with an Intel (Santa Clara, USA) processor. The images were scanned using a 300 dots per inch (DPI) flatbed charged coupled device (CCD) image scanner (model XRS CX3 by X-Ray Scanner Corporation, Torrance, USA) with a 21 × 35 cm scanning area, giving a maximum of a 2478 × 4130 pixel matrix. In practice the image digitization matrix sizes varied from 512 × 512 to 2048 × 2048 pixels depending on the selected resolution and image size. Each pixel had a 256-step (8 bits) grey scale. Large images (e.g. 35 cm × 35 cm films) were scanned in two parts and connected together. A commercial image-processing program was used for scanning and viewing (PhotoStyler by Aldus Corporation, Seattle, USA). The computer screen was 47 cm in diameter. Its resolution was 1024 × 768 pixels, but the image zooming and scrolling functions in the viewing program made it possible to interpret larger digitized images. The image data were transmitted via a dial-up digital 64 kilobits per second (kbit/s) one-channel ISDN telephone line with the help of a communication program (TeleImage by Telecom Finland). The procedure of sending the images consisted of the following steps: (1) selecting the images, (2) prescanning and setting up parameters for the first image, (3) scanning, (4) saving the files, (5) making a connection, (6) sending the images. Image receiving was automatic. Both the on-line sending time and the total image transmission time were recorded. The PC-based teleradiology system was tested in three different system configurations (phases). The software interface and the use of compression differed in these three phases. In the first and second phases, all the programs had to be used separately and the scanning parameters were adjusted manually. No common user interface was used. In the first and second phases, all the images were sent uncompressed. Before the actual clinical cases, a small pilot series of 10 images was made to determine appropriate settings for image digitalization. In the third phase, the effect of data compression and a new easier graphical user interface (programmed with Microsoft Visual Basic) on the total transfer time was tested. All the programs operated in a Microsoft Windows environment. No more than five buttons were available at each screen. Presets for the scanning 45 parameters were used for different images. These presets included a possibility to decrease excessive contrast in chest and bone images. Image resolution for chest and bone images was now set to the same level as in the commercial image plate system Fuji AC-1 (Fuji Corporation, Tokyo, Japan). In this series, images were sent using lossy DCT (discrete cosine transform) image compression with approximately 1:10 compression results (Joint Photographer Experts Group, JPEG compression). (In literature, DCT compression methods with slight compression between 1:10 and 1: 25 depending on the quality of the original have given acceptable results with no statistical difference in diagnostic quality (Lo & Huang 1986, Ishigaki et al. 1990, MacMahon et al. 1991). JPEG compression has been proposed for use in PACS environments (Kajiwara 1992). The requests and reports were written with the Windows Write program (Microsoft Corporation, Seattle, USA) in a separate program window on the same terminal. A total of 254 films were sent, all to a university hospital. In the first phase 48 uncompressed images were sent from a rural health centre, and in the second phase 126 sets of uncompressed images were sent from a central hospital. In the third phase, 80 compressed images were sent from a regional hospital. The selection of patients was made so as to simulate as accurately as possible real teleconsultation needs when radiological expertise is not locally available. At the health centre the images were scanned and sent by a computer operator, while in the central hospital and the regional hospital this was done by x-ray technicians. At the university hospital, the images were interpreted by two staff radiologists on a computer screen. They first scored the images as acceptable or unacceptable in the technical respect. After that, a consensus diagnosis was made from the technically acceptable images. A report was given and returned to the sender via telefax in phases 1 and 2 and as a computer file in phase 3. The same radiologists inspected the original images two weeks later and again made a consensus diagnosis. The diagnoses were compared for possible differences. 4.2 Work practise analysis The study The work practice study consisting of fieldwork (from October of 1995 to January 1996) and workshops (March and May of 1996) was carried out in connection with the implementation and clinical use of the new teleradiology system, which 46 was designed to answer many shortcomings found in the first paper. Especially a new 12 bit gray scale full film area CCD-scanner (Pinja by Finnelpro, Tampere, Finland), a new archiver and communication software (Medilink by Acta Systems, Oulu, Finland) and a new networked teleradiology review workstation with a high-resolution 1600 x 1280 pixel screen and software developed for the purpose (Pinjaview by Finnelpro, Tampere, Finland) with an adjacent electronic text messaging service for referrals and reports were integrated together. The database combining images and text made it possible to follow the status of consultation tasks. For an overview of the teleradiology system configuration, see figure 2. Fig. 2. Teleradiology set-up between Kuusamo primary care health centre and Oulu University hospital. Technical details are explained in the original paper (Paper II, published by permission of Elsevier). The common purpose was to achieve a high usability with an iterative development process after the initial launch of the system. The experimental stage lasted from October 1995 to April 1997. During this period, a total of 375 image packages consisting of several images were transferred. Parallel to the fieldwork, user questionaires were collected of 108 consecutive cases, with questions asking the case distribution and the image processing tools used by the physicians. In the study the reseachers from the Department of Information Processing Science took 47 care of conducting the fieldwork and workshops while the researchers from the Department of Diagnostic Radiology took care of the development and use of the teleradiology system itself, the user questionaires, and expert participation at the workshops. Both teams then contributed to the multidisciplinary presentation of the results with an emphasis on image interpretation. The fieldwork Interviews with design team members and participants on the use of the teleradiology system were conducted during the fieldwork (Karasti 1997) with total of 12 interviews. Emphasis was, however, on studying the actually occurring clinical practice of teleradiology in everyday settings in both locations (approximately 15 working days) by observation, contextual interviews and videotaping. A total of 21 system use-related instances of work were recorded (11 image interpretation sessions, 4 instances of typing out reports and sending them, 6 instances of scanning and sending images). Stimulated recall interviews (Engeström 1996) were carried out with the radiologists whose interpretation work had been recorded (7 patient cases). The interviewees and researchers watched the tapes together, questioning and commenting on the events recorded. Analysis of work practice started already during the fieldwork as the fieldworkers’ interpretations and formulations of the phenomena were continuously challenged by new observations. The stimulated recall interviews also served as collaborative analyses. After the fieldwork more focused and detailed video analyses were conducted with special attention on the unfolding of work: interaction and use of materials, skills and knowledge employed in work. The workshops Based on the fieldwork and the analyses of work two workshops were organized (Karasti 1997). The communities of work practice, design and research were brought together to collaboratively evaluate and redesign the system so as to support the established and contingent practices through which the work is accomplished and to address issues relevant to developing the teleradiology work practice. Participation of all occupational groups involved in the teleradiology practice was required. A video collage of actual work situations was edited to present the entire teleradiology work practice in parallel to corresponding traditional ways of working. The video collage served as a shared object of 48 interest and helped provide a basis for the cooperative analysis of actual work practice. The participants were able to compare practices and share their experiences. They built a common understanding of the teleradiology work practice. System usability was evaluated based on cooperatively produced emergent practical criteria. Moving on from evaluation to envisioning future use situations they started to raise informed issues relevant to both system redesign and development of work procedures. 4.3 Internet teleradiology An experimental consultation system was tested between Oulu University Central Hospital in Finland, University Hospital of Reykjavik in Iceland and University Hospital of Tromsø in Norway. The hospitals provided a gateway access (firewall policy) to the NORDUnet network and workstations for image viewing and interpretation. Access to digitized images occurred either by network connection to image sources or by film digitization with a CCD scanner. Digitized images were converted to common TIFF (tagged image file format) or JIF (JPEG interchange format) file formats and sent using Internet FTP (file transfer protocol) protocol. Technical details of the hardware are summarized in Table 2. Table 2. Technical details of the hardware used in multicentre Internet teleradiology (Paper III, published by permission of Springer). Hardware Reykjavik Digitizer CCD CCD CCD CCD Pixel size 85 µm 170 µm 85 µn 85 µm Grey scale 8 bits 12 bits2 8 bits 8 bits 3 Gateway Unix/HP Viewing station PC 486 Display pixels 1280x1024 1 Unix/Sun 4 Tromsø 1 Tromsø 21 Oulu components Unix/HP/Sun Unix/HP PC 486 Unix/HP/Sun Unix/HP 1280x1024 1280x1024 1280x865 Tromsø used two different hardware constellations, 2 only for digitalization, image display grey scale was always 8 bits. 3 HP is a trade mark of Hewlett-Packard. 4 Sun is a trade mark of Sun Microsystems. In the first study, 131 randomly selected neuroradiological CT and MRI images plus skeletal x-rays were sent between Oulu and Reykjavik. Different compression methods (lossless LZW (Lempel–Ziv–Welch) with compression up to 1:2.4 and lossy JPEG compression up to 1:10) and image resolutions were 49 tested and transmission times recorded. The images were compared with originals for possible differences in quality or diagnosis. In the second study MRI studies of pituitary glands were sent between each institution. Laser-printed images were digitized and compressed according to the results of a previous study with lossy JPEG-compression with quality factor 85, giving approximately 1:10 compression. Twenty pituitary cases were sent from Iceland both to Norway and Finland. In the same manner, 20 cases were sent from Finland to Iceland and Norway. These four sets made a total of 80 cases. The clinical information given was always the same: suspected enlargement of sellar contents or suspicion of microadenoma. In every institution three neuroradiologists interpreted the digital images on the screen and indicated each abnormality according to a 5-point confidence scale: 1= definitely not present, 2=probably not present, 3=cannot be decided, 4=probably present, 5=definitely present. The interpreters used grey-scale windowing and zooming functions on the workstation. The original analogue radiographs were interpreted in the same manner two weeks after the workstation session. The actual diagnosis was determined by consensus of two neuroradiologists in the sending site using all the patient information available. Receiver operating characteristic (ROC) curves were produced and the area below them and statistics calculated by using methods described in radiological literature (Hanley & McNeil 1982, Hanley & McNeil 1983). In connection with this study 12 clinical consultation cases were sent from Iceland to Finland, recording the time of sending, viewing and reporting. These cases included both images and the request for examination. For data safety reasons, patient identification data were removed from all the material sent. 4.4 Mobile PC teleradiology Portable terminal The neuroradiologist was supplied with a portable terminal capable of receiving and viewing images. The terminal was constructed from a notebook computer (Contura 430cx, Compaq Computer Corp., USA) equipped with an active matrix display of 640 × 480 pixel resolution. The display was capable of showing 256 colours. The computer weighed 1.7 kg, measured 23 × 30 cm and cost about EUR 50 3150. The computer had a 16 MByte random access memory and an 800 MByte hard disk. MS Windows (Microsoft Corp, Seattle, USA) was used as the operating system. Images were viewed with a shareware program (Polyview, Polybytes, Cedar Rapids, USA). The notebook computer was equipped with a PCMCIA digital cellular data card (Nokia Mobile Phones, Salo, Finland). This was used to interface the computer to a GSM telephone (Nokia 2110i, Nokia Mobile Phones, Salo, Finland). When the neuroradiologist was not in the hospital, the data connection to the hospital was established through the GSM network. A commercial GSM data service (Telecom Finland, Helsinki, Finland) was used for the wireless connection. The nominal data transmission rate was 9.6 kbit/s. The telecommunication software used was the Microsoft remote access service of Windows 95. The remote access service included PPP (point-to-point protocol) client software that could be used to connect to a PPP server at the hospital over a modem line. Microsoft’s implementation of the TCP/IP was used over the PPP connection. PPP is the certified Internet standard protocol that extends the LAN protocols to function over a modem line. When the TCP/IP connection was established, a data transmission connection was formed and the images were transferred using an FTP program (WS-FTP, Ipswitch Inc, Lexington, USA). Dial-up server To provide a dial-in connection to the LAN of the Department of Radiology, a PPP server was set up. The server accepted incoming modem calls and, after user authentication, provided a connection to the LAN. When the connection was active the server routed network packets between the modem line and the LAN. In the present study, the server was used to provide a TCP/IP connection between the LAN and the portable terminal. Initially the PPP server was based on a Unix workstation (Indigo, Silicon Graphics, Mountain View, USA). Subsequently, it was changed to a standard Intel Pentium (Santa Clara, USA) personal computer. The computer used Linux, a freely distributable implementation of Unix, as the operating system. 51 Imaging device interface The patients were imaged with a CT scanner (GE HiSpeed Advantage or GE Sytec 3000, GE, Milwaukee, USA). The images were routinely transferred to a visualization workstation (SUN Microsystems, Mountain View, USA) via the LAN. The visualization workstation used a diagnostic visualization program (Advantage Windows, GEMS, Buc, France). The visualization workstation had a display with 8 bit/pixel colour depth, which was able to show 256 colours simultaneously. The contrast window and the brightness level of the images were set to make the images suitable for diagnosing from the monitor. For the technical set-up, see figure 3. Fig. 3. The technical configuration of the wireless image consultation chain. The portable terminal connects to a private dial-up server for a secure image consultation (Paper IV, published by permission of RSM press). After manipulation, the images were captured from the screen using public domain xv software with a special script. This was performed by an X-ray technician who had detailed instructions to manipulate the images to look the same as if printed on film. The capturing took on average 7 min per patient. The colour depth used was the full dynamic range of the display, i.e. 8 bits/pixel. The images were captured at the original 512 × 512 pixel resolution. The captured image size was 256 kBytes in size. All the captured images from the visualization workstation were stored in a network directory that was physically mounted on the dial-up server. 52 To reduce the bandwidth required for transmission, the images were compressed with a JPEG algorithm and stored in 8 bits/pixel greyscale format. The JPEG quality factor used was 75%, resulting in an approximate compression ratio of 8:1. This yielded an average image size of 40 kBytes. The images were then automatically transferred in patient folders to the dial-in server. Patients and interpretation The material consisted of 68 consecutive neuroradiological emergency patients (36 males and 32 females). The mean age was 59 years, range 1–90. There were 63 head, 1 lumbar spine, 2 sinus, 1 orbital and 1 neck CT examinations. Two series of scans were examined. In the first series, a neuroradiologist interpreted the CT scans of 47 patients on a portable computer during working hours. Later, the same neuroradiologist reviewed the images on film. In the first series, all the CT slices obtained in each examination were transmitted. The experience was used to define the second part of the study. In the second series, the images of 21 patients were sent by a junior radiologist for a neuroradiologist consultation after normal working hours. Three senior neuroradiologists gave their advice in turn. These results were also compared to the film interpretation the next morning. In the second series, the junior radiologist either sent all slices or only those slices covering the area requiring consultation. In the re-evaluation, the images of all patients were viewed in the manner used by the radiologists to interpret CT-images in daily work. A written report was made for both transmitted and comparison images. The images of patients were transferred to the portable terminal by GSM. The image transfer times (both FTP connection time and total time) were recorded. After receiving the images, the radiologists evaluated them for diagnostic quality, scoring them as: 1=suitable for final diagnosis, 2=suitable for preliminary diagnosis only or, 3=not suitable for diagnosis. The reasons for inadequacy were recorded. The neuroradiologist also evaluated each image consultation on a fivepoint scale indicating whether the image consultation was beneficial compared to an ordinary telephone discussion. The scale was: 1= no need for consultation, 2=no benefits over a telephone conversation, 3=minor benefit, 4=beneficial for diagnosis, 5=gave essential information. The neuroradiologst also decided and recorded whether a hospital visit would have been necessary without the image transfer. 53 4.5 Smartphone teleradiology 4.5.1 Feasibility study PDA terminal The PDA terminal device for image reception and viewing was a Nokia 9000 Communicator (Nokia Ltd, Helsinki, Finland), which was a combined digital phone and messaging device with short message service (SMS), telefax and even Internet/intranet capability. It could be called an intelligent phone (smartphone) or personal digital assistant with GSM phone capabilities. It weighted 397g and its dimensions were 173 × 64 × 38 mm. The LCD (Liquid Crystal Display) display area was capable of showing graphics of 600 × 200 pixels with 8 grey levels, and the contrast was adjustable. Larger images could be viewed using zooming and scrolling functions. It had 8 megabytes (MB) total memory, 4 MB for operating system and applications, 2 MB for program execution and 2 MB for user data storage. The cost of the terminal device was 1,000 Euros. As an extra capability, the phone was equipped with Celesta MDO Mobile Data Organiser (CCC Mobile Ltd, Oulunsalo, Finland) file transfer software. Hospital teleradiology server and image acquisition To provide a dial-in connection to the LAN of the Department of Diagnostic Radiology, a private PPP server using a Microsoft remote access service (RAS) of Windows NT (Microsoft Corp., Seattle, USA) was set up. PAP (Password Authentication Protocol) was used as an authentication mechanism for PPP connection. After user authentication a TCP/IP connection between the PDA terminal and teleradiology server of the hospital was established using commercial GSM data service. Internet standard protocol FTP was used for transferring the images from the teleradiology server to the PDA terminal. The patients were imaged with a CT scanner (CT HiSpeed Advantage or CT Sytec 3000, GE, Milwaukee, USA) and the image capture arrangement was the same as reported in the previous work on mobile PC teleradiology and described in the previous chapter. 54 Patients and interpretation A junior radiologist at the hospital selected either all the CT slices or slices containing the area requiring consultation to be stored in the teleradiology server. After notification of an emergency case, a senior neuroradiologist at home made a data call, logged into the server and downloaded and viewed each of the 21 patient cases. Twenty of these were head scans and one was a sinus scan. There were 12 female and 9 male patients, their mean age being 62 years (range 1–87). Image transmission, file handling and image interpretation times were recorded as well as any technical remarks. The neuroradiologist rated the images for image quality; the scoring was as follows: 1. suitable for final diagnosis; 2. suitable for preliminary diagnosis only, or 3. not suitable for diagnosis. The reasons for inadequacy were recorded. After that the neuroradiologist gave a written report. These reports were compared to the original report given at the hospital and to a reference reading from the original films the following day. The neuroradiologist also scored each image consultation on whether it was beneficial compared to a telephone discussion. The scale was: 1. no need for consultation; 2. no benefits over a telephone conversation; 3. minor benefit; 4. beneficial for diagnosis; 5. gave essential information. The neuroradiologist also recorded whether a hospital visit would have been necessary without the image transfer. 4.5.2 Clinical study MOMEDA According to the concept, a Mobile Medical Data server was established in the hospital intranet. In the case of consultation, the server received the DICOM images and sent them together with relevant narratives retrieved from the hospital EPR system. After an automatic SMS initiation, data connection was established with a secure call-back method and connections were authenticated with a double security verification (both at the PPP server and at the communication software). The terminal device was a Nokia 9110 communicator device (Nokia Ltd, Helsinki, Finland) which was smaller (158 × 56 × 27 millimetres) and more lightweight (253 grams) than the original model, yet having still monochromatic display with 640 × 200 pixels and data transfer with GSM data technology (PDAdb.net 2007b). The developed client image viewer software (Celesta Ltd, Oulu Finland), was a light-weight version of diagnostic image workstation viewer software, and was 55 capable of full diagnostic manipulation of data. Additional information requests could be made from the hospital EPR system using a WWW-browser client over a secure connection. Finally, a consultation phone call could be made to the hospital with an embedded hands-free phone while reading the images. For the mobile terminal functions, see figure 4. Secure Mobile Terminal Access after Local User Authentication SMS-Receiver/ Listener Secure Call-back with Authenticated Server Access and Image Transfer Radiology Workflow Functions RIS Worklist / Patient Information Referral Text Display Image Overview/ Thumbnail Display Error Recovery EPR Functions WWW-Browser/ EPR Query Image and Text Unpackager Communication Hands-free Phone EPR Display EPR Text Input Telefax Full Image Display Image Processing/ Window/Level, Zoom, Scroll, Flip, Turn Basic Office Tools: Notetaker, Spreadsheet, Calender, Calculator, Clock Fig. 4. The radiology workflow and security oriented software components realized in the integrated mobile image and electronic patient record terminal. PROMODAS The Professional Mobile Data Systems project developed the ideas further to 2.5 generation wireless networks and platforms. GPRS technology with IP-protocol facilitated the fast development of the system because the amount of devicespecific programming decreased. Improved image manipulation functions were developed for the mobile client software. This first Microsoft Pocket-PC platform smartphone device O2 XDA (HTC Corporation, Taiwan) was even more lightweight with 201 grams, yet the processor was more powerful (PDAdb.net 2005). The colour display (240 × 320 pixels, 4096 colours) in this combined PDA and GPRS mobile phone was larger in vertical direction and capable of showing more grey levels. The security of the data transfer was also improved, and VPN tunnelling with strong encryption was utilized. The system architecture is shown in figure 5. 56 Fig. 5. Schematic architecture of the PROMODAS system (Paper VI, published by permission of Elsevier). Evaluation The clinical feasibility of the MOMEDA and PROMODAS systems was tested in two different evaluations with real CT and MRI on-duty scans sent to the mobile terminals. In the first clinical evaluation, 115 CT and MRI scan series were read with the MOMEDA terminal. In the second evaluation 150 CT and MRI scan series were read with the PROMODAS terminal. Before the evaluation phase a short educational pilot was conducted. The reading radiologists, neuroradiologists and neurosurgeons also had the full examination request coming from the EPR available at the mobile terminal and were able to use image processing tools at each of the mobile terminals. Detailed reports were made of each examination, and after a time interval compared to reports made using the full-size workstations at the department of radiology. In addition to the results of image interpretation also the amount of time needed for image transmission and reviewing the cases was recorded. 4.6 Integration of teleradiology as a part of eHealth services The structured web-based questionnaire of eHealth usage was distributed by email to all public health service providers or hospital districts and health care centres, and to a sampling of private health care providers. The survey was conducted for the first time by FinnTelemedicum in 2003 (Kiviaho et al. 2004) and repeated in 2005. In this study, data from the year 2005 are presented with comparison to data from the year 2003 where appropriate. The questionnaire 57 consisted of: the identification of the responding organisation and the respondent; questions about the adaptation of EPR systems; systems or applications used to transfer/exchange patient information between organizations during care processes and the standards in use for the migration of patient information; methods of authentication, identification, and informed consent of patients; the usage of different eEducation systems for staff education; the types of human and material resources needed; systems supporting quality control and service delivery; and the adaptation of different eServices for patients. The total number of questions was 97. Most of them also included further questions on how old the system or application concerned was and the intensity of use. The questions for hospitals, health care centres and private health care providers differed to some extent, depending on the nature of the services they provided. The intensity of use told about the amount (%) of the action or function being carried out by electronic means. For example, if a service provider used EPR for the documentation of patient data in half of the cases and a paper-based record for others, the intensity of use of the EPR was 50%. Several of the questions in the year 2005 survey were the same as the questions in the FinnTelemedicum survey conducted in late 2003, using the same web-based data collection methods. The questionnaire was emailed between October and November 2005 to all public service providers, i.e. to 21 hospital districts and 251 health care centres. The questionnaire was also emailed to a sample of 65 private health care service providers including 30 of the largest service providers, and 35 who had responded to the 2003 survey. Additional information was obtained by telephone interviews from those health care sites where not all information in the returned questionnaires was complete. Also PACS vendors were interviewed for their reference sites in primary care. A full report with a detailed description of the method has been published in Finnish (Winblad et al. 2006) and an English version with edited information has been published by Hämäläinen et al (2007). 58 5 Results 5.1 Low-end technology teleradiology Time factor The average total image transfer procedure time in phase one (from the health centre) was 15 min 50 sec (range 15 min to 18 min 20 sec) for an uncompressed single chest x-ray image (average size 3897 bytes). The longest single task in the transmission process was the on-line sending of image (average 10 min 50 sec), followed by scanning (64 sec, range 55–74 sec) and saving the images on a hard disk (57 sec, range 53 to 83 sec). The average throughput of the digital line was 60 kbit/s (range 56 to 61 kbit/s) per second. The average total image transfer procedure time for a CT film (4 to 6 images on each film) with an average size of 840 kbytes in phase two (from the central hospital) was 6 min 20 sec (range 5 min 10 sec to 7 min 20 sec) per uncompressed film. The longest single step was sending the film (2 min 20 sec), the next longest step was scanning (35 sec, range 23–42 sec), followed by saving the films on a hard disk (26 sec, range 19 to 35 sec). The average total image transfer procedure time for a 1:10 compressed chest x-ray film in phase three (from the regional hospital) was 5 min for a single view (36–49% of the time for uncompressed images). Of this, only 30 sec on average was on-line time (5–9% of that for uncompressed images). The average time required for the preparatory steps preceding the scanning had decreased from 3 min to 2 min 30 sec. Technical quality Of the 48 uncompressed films transmitted in phase one (from the health centre), two chest x-rays and five bone images were considered technically unacceptable. This was due to faulty scanner settings; one chest x-ray had an excessively small scanning area and the other was initially digitized too light. Two thoracic and three cervical spine images were scanned with excessive contrast settings, with details lost in both light and dark areas. All the uncompressed transmitted images in phase two (from the central hospital) were considered technically high-quality. 59 Of the 80 compressed films sent in phase three (from the regional hospital), the only low-quality film was an image of the thoracic spine, which was digitized with excessive contrast settings, with loss of detail in the highlight parts as a consequence. Diagnostic ability Diagnostic agreement between the transmitted and original images was determined on all of the technically acceptable chest (n=8) and bone (n=33) images transmitted uncompressed in phase one (from the health centre). A lesion was missed in two cases of the 126 image sets (2%) transmitted uncompressed in phase two (from the central hospital). Both cases were brain CT scans (n=55): The first was a small infarction in the cerebral pons (diameter under 5mm), while the second was a thickening of the right nervus opticus (difference of 5 mm compared to the other side). All the 38 body CT sets as well as all the 25 angiography film sets, 4 mammographies and 4 ultrasound images were correctly interpreted from the transmitted images. A discrepant finding was made in four (5%) out of 79 image sets satisfactorily transmitted in compressed form in phase three (from the regional hospital). All of them were mammography images (n=40). After interpreting the original images, an enlargement mammography was considered necessary in two cases, while in two cases no further examinations were considered necessary even though an enlargement mammography was suggested after reading the transmitted images. All the twenty chest images and all the nineteen bone images were interpreted correctly from the transmissions. Total diagnostic agreement between the transmitted images and the original films was 98% in the technically acceptable image transmissions. 5.2 Work practice analysis Moulding two previously separate communities to work together From the point of view of the Kuusamo Primary Health Care Centre, teleradiology became a question of integrating and scheduling expert consultation into the other constituents of patients’ care trajectory (Strauss et al. 1985). The two previously separate communities of radiological work practice began to 60 cooperate in a distributed fashion. They had to revise their local ways of working and establish new commonly agreed work procedures. The teleradiology system replaced the peripheral awareness and face-to-face communication of local work communities by asynchronous computer-mediated image transfer, request-report exchange, and occasional phone calls on critical cases. Interactive cooperation was replaced by distribution of tasks performed in geographically separate locations at sequentially ordered times. The new teleradiology work practice is depicted in figure 6. Fig. 6. The teleradiology work practice between primary care and secondary care consultant centre. (Co-author Helena Karasti is thanked for designing this figure. Paper II, published by permission of Elsevier.) The chain of tasks After radiological examination of a patient and initial interpretation of images, the attending physician in Kuusamo decides whether an expert consultation is necessary, in which case he or she dictates a request for a consult. An X-ray technician or secretary scans the films, enters the patient information, types the request into the system and then sends them to the University Hospital. The teleradiologist on duty at Oulu University Hospital checks once a day whether images to be interpreted have been received. The radiologist then interprets the 61 images of each patient case on a computer screen and dictates reports. She or he hands over the tape to the departmental secretary who types the report into the system and sends the file back to Kuusamo Primary Health Care Centre. The reports are received by an X-ray technician who prints them out to be delivered to the attending physicians. Activities aimed at ensuring a smooth flow of work A great deal of work was carried out by the personnel in addition to the chain of tasks directly related to using the teleradiology system. A continuous stream of contingent support work was needed in “making the system work” (Bowers et al. 1995) as the patient case was taken through the consultation service. This articulation work (Strauss et al. 1985) was carried out both in individual tasks and on the level of cooperating the entire process. For example, prior to the image transfer, the X-ray technician on duty prepares the patient folder. Films of the patient’s previous radiology examinations are acquired, relevant films selected, any missing information is checked and the patient material is organized. The technician carries out all these measures to ensure a smooth flow of work in the system while scanning, entering and sending information. Another set of examples is of activities carried out to sustain an account of the service as a formal sequential operation. The participants routinely checked whether they had a task to carry out and after “completing their bit” made sure the next step was completed, i.e., a sender in Kuusamo made a phone call to Oulu to inform about a critical case, a radiologist handed over the dictation tape to a secretary, who in turn, after typing out and sending the report, checked whether it was received at the Kuusamo end. The changed script of image interpretation work In the film-based image reading practice, smooth flow of work for image interpretation is ensured by a great deal of support work performed by other occupations and easy-to-use technology in the form of films on autoalternator light screens. Radiologists are accustomed to interpreting images as sequentially ordered processes of cases in series. The alternator supports this with its magazine for light screens that are used for the temporary storage of organized films. In traditional image interpretation sessions, the radiologists are accustomed to 62 reading the request concerning the following patient to determine the problem formulation while the light screens are swiftly changed between the cases. As the next patient’s pair of light screens appears, they take a quick overall glance at the films and then proceed to a more detailed interpretation process and dictation of a report. In a similar fashion, in a typical teleradiology image interpretation session, a report was first displayed on the screen followed by images. The radiologist would, however, have to wait idly while the images appeared slowly on the screen one by one in blown-out-size and were then reduced to fit the pre-set frame for each image. Then the radiologist would have to reorganize and resize the images to suit the more detailed interpretation process. The system caused disruptions and breakdowns in the smooth organization of work by imposing a new script for image interpretation. Dispersion of the patient records and the examination process The radiologists are used to having a patient folder easily accessible on the alternator’s desk during image interpretation. Should a need for more information arise it can be referred to, at least for images of previous examinations with accompanying reports and on occasion also for the patient’s medical record. Material transferred via teleradiology from the primary health care centre is less comprehensive than radiologists are accustomed to, being minimal (“it’s like the patient was to a radiological examination for the first time”, “there is no history”) or not sufficient (occasionally more images had to be requested to enable comparison). Dispersion of the examination process led to situations where the teleradiologists did not have access to knowledge that is self-evident when the examination is performed within close proximity of a department. They may miss the locally produced knowledge related to the patient’s examination, patientrelated information (the patient’s physical appearance, conduct and behaviour, and answers to the questions posed during examinations), and even recollections of the patient’s previous examinations. Dispersion of image articulation work The radiologists need to be able to concentrate on the image interpretation work during the sessions. To ensure an efficient flow of image interpretation, all imagerelated articulation work is carried out prior to the session. In the film-based work 63 practice at the Oulu University Hospital, the retrieval, checking, organization, arrangement and layout of films on the alternator’s light screens was done by a special assistant. With the studied new teleradiology system, the tasks corresponding to the assistants’ work were dispersed into four aspects of the teleradiology system-related work practice: 1) the scanning of films and transmission of images in Kuusamo – which affects the final display order and orientation, 2) the image display system – how it displays a series of images on the computer screen, 3) how the radiologist arranges the images on the computer screen at the beginning of a session and 4) the continuous image organization and processing by the radiologist during the interpretation process. In the teleradiology experiment, the radiologists ended up doing a great deal of image-related articulation work both at the beginning of each case and during the interpretation process itself. One factor contributing to the latter are the enhanced image processing functions. During the study they were used as follows: zoom function (enlargement) in 71% of the cases, image windowing and levelling (brightness contrast adjustment) in 70% of the cases, image rotation (correct side up) in 25% of the cases and image enhancement (unsharp mask function) in 3% of the cases. Another contributing factor is the lack of system support for an easy way of laying out images for juxtaposition on the screen. The radiologists’ part of the system was designed based on the assumption that the transferred cases would mostly be acute ones with few images. It turned out that 25% of the cases were acute ones (report needed in less than an hour) while the rest were urgent ones (75%, report needed in less than 24 hours). However, many of these were control cases with numerous images (19% of the total number of cases), and even in the acute cases comparison of images side-by-side was central to image interpretation. To a great extent, the professional expertise of radiologists lies in being able to follow a change through a series of images. This is traditionally done by comparing films side-by-side on large light screens. 5.3 Internet teleradiology Image quality, transmission time and interpretation time In the first phase, image quality was found to be adequate with both uncompressed and compressed images. Network transfers occurred without errors or data loss. Data conversion between MS-DOS and Unix computer platforms 64 also occurred without problems. Image transfer times between Iceland and Finland ranged from 2.5 minutes (compressed JIF file size 140 kbytes) to 43 minutes (uncompressed TIFF file size 2480 kbytes). In the second phase, where only JPEG compressed images were sent, the average image transfer time between Finland and Iceland was 3 minutes 45 seconds for an average image size of 270 kbytes. The average image transfer time between Finland and Norway was 55 seconds for an average image size of 317 kbytes. As each study case needed two to three images, total on-line transfer times ranged from 7.5 to 11 minutes (Finland-Iceland) and from one to three minutes (Finland-Norway). The average radiologist’s working time at the receiving site for the clinical consultation cases was 8.5 min (range 5 to 14 min). This covered time from the opening of images on screen to the end of the dictation session. Comparison of screen and film readings in sellar pathology The results of each group of three neuroradiologists and their combined reading results in the second phase as analysed by the ROC study are summarized in Table 3 and combined ROC curves are presented graphically in figures 7 and 8. Table 3. Area under ROC curve derived from the analysis of the enlargement of sellar contents and from the analysis of possible presence of microadenoma (III, published by permission of Springer). Evaluation of sellar enlargement Evaluation of presence of microadenoma Area under ROC curve Area under ROC curve Reader group Screen Film Reader group Screen Film Group 1 ISFI 0.874 0.869 Group 1 ISFI 0.762 0.803 Group 2 ISNO 0.884 0.933 Group 2 ISNO 0.755 0.823 Group 3 FINO 0.799 0.820 Group 3 FINO 0.613 0.590 Group 4 FIIS 0.848 0.848 Group 4 FIIS 0.758 0.738 Total 1-4 0.842 0.863 Total 1-4 0.758 0.717 ISFI = Sending Iceland -> Finland, ISNO=Sending Iceland -> Norway, FINO = Sending Finland ->Norway, FIIS = Sending Finland ->Iceland. There were no significant differences in readers’ performance between teleradiology screen readings and film readings of the images either in each individual reading group or in combined reading results for each study question. 65 Fig. 7. The combined ROC curves of all the readers evaluating the possible enlargement of sellar contents. The area under the curve for teleradiology screen reading is 0.842 and for reference film reading 0.863 (III, published by permission of Springer). Fig. 8. The combined ROC curves of all the readers evaluating the presence of pituitary microadenoma. The area under the curve for teleradiology screen reading is 0.719 and for reference film reading 0.717 (III, published by permission of Springer). 5.4 Mobile PC teleradiology In the final evaluation of the 68 consecutive neuroradiological emergency patients, there were 22 normal cases and 46 pathological cases. The most common 66 pathological findings were brain infarction (16), haemorrhage (12) and atrophy (7). In all cases the consultations succeeded technically. The average data transfer time (i.e. the FTP connection time) for a single compressed CT slice via GSM was 55 s (total connection time one min). The transfer of an ordinary head CT examination took 18 min on average. The transmitted images were considered suitable for final diagnosis in 49 of the 68 cases (72%) and suitable for preliminary diagnosis only in 19 of them (28%). None of them were considered not suitable for diagnosis. The reasons for suitability for preliminary diagnosis only were: the need to use a different window and level (n=10), the need to compare to previous studies (n=7), or the need to see more images than those selected for consultation by the junior radiologist (n=2). The diagnosis from the transmitted images was the same as that from the original images in 66 cases (97%) and slightly different (although no effect on patient care) in two (3%). A small brain infarction in the posterior fossa region remained undiagnosed and an artefact 2 mm in diameter was misdiagnosed as a pons infarction. Compared to an ordinary telephone discussion, the transmitted images gave essential information in 55 cases (81%), beneficial information in eight cases (12%) and minor benefit in five cases (7%). According to the consulting radiologist’s opinion, the mobile image consultation saved him a hospital visit for image interpretation in 16 cases (24%). 5.5 Smartphone teleradiology 5.5.1 Feasibility study The 21 patient cases contained on average 14 images (range 3–20 images) per case. The distribution of the diagnoses in original images was: 11 brain infarctions, 2 atrophies, 4 haemorrhages, 1 aneurysm, 1 infection and 2 tumours. Connection to the server was in all cases established with the first attempt. The connection initializing time was on average 1 min 20 s (range 1 to 2 min). In three cases the data connection was lost before all the images were transferred, but a new connection was established after notification of an error message. There was on average 2 min 40 s time (range 1 min to 6 min 50 s) to select the correct images from the server. Image transmission to the terminal took 1 min 30 s per 67 image (including image transmission and file management time). The average net transmission time for a head CT-scan was thus 21 min (range 4 min 30 s to 30 min). After the connection was closed the local interpretation of the images took on average 15 min (range 7 to 23 min); the mean total working time on one case was thus 40 min. There were no changes in the data contents of the transferred data. In one of the cases the image quality was good enough for a “final report” and in 20 cases the quality was suitable for a “preliminary report”. There were no cases in the category “not suitable for diagnosis”. The reasons for categorizing images suitable for “preliminary report” were neuroradiologist’s doubts about the suitability of image quality for interpreting small changes in 18 cases and need to window and level the grey scale of the images in 2 cases. The report given from the transmitted images was identical with the reference film reading in 18 cases (86%) and not totally identical in 3 cases (14%). The differences were: 1) a possible density loss at nucleus lentiformis was not seen, 2) a small lacuna infarction at the right capsula interna region was not found, 3) a pineal cyst was not found. In one case a pons infarction observed on the PDA terminal by the neuroradiologist had not been noticed by a junior radiologist in the original film report. According to the neuroradiologist, the transmitted images gave essential information in 5 cases (24%), beneficial information in 13 cases (62%) and minor benefit in three cases (14%) compared to an ordinary telephone discussion. The neuroradiologist considered that the image consultation saved a hospital visit in 15 cases (71%). 5.5.2 Clinical study MOMEDA With the second generation MOMEDA GSM terminal the sending time of a CT scan series over the GSM network ranged from 15 to 20 min. The reading time for a series of CT scans was 20 to 30 min for radiologists and less than 10 min for neurosurgeons. The reading time for MRI scan series was longer, more than 35 min. This is mainly explained by the larger number of images in MRI examination of the head compared to head CT. The image quality was sufficient for final diagnosis in 38% of the 115 cases and for preliminary consultation in 68 62% of the cases. The diagnostic quality was considered adequate and user acceptance for emergency service was high. The costs of the transmission were related to the transfer time. The demonstrator in this study showed that it is medically sensible to use pocket-sized teleradiology terminals for consultation purposes in neurosurgery and with limitations in radiology. This meant that the final product was adopted for clinical use by the neurosurgeons of Oulu University Hospital in June 2000. PROMODAS With the 2.5 generation PROMODAS GPRS terminal the transmission time was cut down to 5 to 10 min depending on the network load. The image viewer application had similar functionalities as the earlier terminal, but there were considerable changes in the responsiveness of the system in favour of the PROMODAS system, so user acceptance was high. New portable terminal devices with faster processors enabled a more efficient processing and display of patient information and the reading times were reduced to 5 to 10 min per scan series. The confidence in image quality was somewhat higher than in the MOMEDA project; now, image quality was considered good enough for final diagnosis in 40% of the 150 cases, and in 60% good enough for preliminary consultation. Telecom operators had now implemented a monthly flat fee work for data, which made it possible to stay connected at a lower cost than previously. In practice, the PROMODAS terminal replaced the earlier generation in clinical use by neurosurgeons in June 2003. The functional differences influencing the performance of MOMEDA and PROMODAS are summarized in Table 4. Table 4. Main functional differences between the two mobile teleradiology systems (Paper VI, published by permission of Elsevier). Parameter MOMEDA Network GSM PROMODAS GPRS Access method Private Internet access point Mobile IP via operator Pricing scheme Per second Flat monthly fee Average download time less than 60 s / image less than 20 s / image Terminal display size 640x200 pixels 320x240 pixels Image decompression time more than 10 s / image less than 1 s /image 69 5.6 Integration of teleradiology as a part of eHealth services Coverage Responses to the year 2005 questionnaire were obtained from all hospital districts (100%, n=21). A total of 179 (71.3%) of 251 health care centres responded to the questionnaire. The responses covered 88.2% of the whole population of Finland. Additional data were also obtained from 27 health care centres that had not responded to the present survey, but had responded to the 2003 survey. Results were obtained from 28 private service providers, about half (n=13) from the biggest private service providers and about half (n=15) from those companies that had responded to the 2003 survey. The private providers included enterprises, from conglomerates with hospital and operative services to small part-time general practices. Because the private providers are a heterogeneous group, the results concerning them can only be regarded as indicative. The health care centres which did not respond were smaller in size than those who did, covering only 12% of the population of the country. Therefore, the answers obtained can be regarded as representative in terms of primary and secondary health care. Because the public sector covers 85% of all health services, the results can be regarded as representative for the whole country. Electronic Patient Record In specialized health care EPR was in use in all but one of the 21 hospital districts. In one hospital district, EPR was in the planning stage at the end of 2005. Among 17 of the 20 users of EPR the intensity of use was over 90%. One hospital district had intensity of use between 50–90%, and two between 25–49% In the year 2003 survey, 13 of the 21 hospital districts had EPR in use. In 2005, 18 hospital districts indicated that the usage of EPR covered at least three out of four main medical responsibility areas (conservative, operative, emergency care and psychiatric care), while only 5 hospital districts had achieved this level in 2003. In primary health care centres, EPR was in use in 240 (95.6%) of the 251 health care centres, three of them had it at testing stage, and eight at a planning stage. In 2003 the coverage was 93.6% in heath centres. The 11 health care centres lacking EPR in 2005 were small and remote, and some of them were planning on converging with a larger neighbouring health care centre. The 70 intensity of use was very high: at 91% of the health care centres using EPR, the rate was over 90%, while among the rest it was 50–90%. Of the 28 responders of the private health care service providers, 25 (89%) used EPR. Compared to the 2003 surveys the figures seem to be similar. The intensity of use was high: three out of four providers had an intensity of use of more than 90%. Picture Archiving and Communication Systems The results of the PACS survey for secondary care (hospital districts) in 2003 and 2005 are presented in Table 5. The progress has been fast during the two years. Fifteen out of 21 hospital districts had a PACS in production in 2005. What is more important is that they were practically filmless (over 90% of the medical images were utilized only in digital form). The rest of the hospitals planned to have their PACS in production towards the end of the year 2007. After that, all the hospital districts in Finland have been totally filmless. Table 5. PACS installations in 21 Finnish hospital districts in 2003 and 2005 (Reponen et al. 2008, published by permission of IOS press). Measure 2003 2005 PACS in production phase 10/21 15/21 PACS in pilot phase 1/21 2/21 PACS in installation phase 10/21 4/21 PACS usage > 90% 6/21 15/21 PACS usage 50–90% 3/21 1/21 PACS usage < 50% 4/21 1/21 Concerning primary care, in the year 2005 full picture archiving and communication systems or system components were in use in 95 out of 179 health care centres (53%), based on the information provided by the vendors. In the previous survey conducted in 2003, PACS usage information was obtained directly from the primary health care centres; at that time, it was reported that PACS components were used in 27 (17%) primary health care centres. Even though the methodology is different, this information reveals that the use of PACS at primary health care level has increased in Finland. 71 Results of regional services Before presenting the results of many different and yet at the same time partially overlapping forms of interorganizational data exchange, a couple of definitions are needed. Electronic referrals are basically sent to another institution in order to transfer the responsibility of patient care. Electronic discharge letters are then returned to the sending institution once the patient’s treatment is finished. The referral can evolve to an electronic consultation letter, if neither responsibility for the patient nor the actual patient is transferred, but professional advice for treatment is sought or professional opinions are given. Consultations can also take place with a help of videoconferencing (televideo consultations). There are special cases like teleradiology, which can be used for consultation but also for information distribution; the same also applies to telelaboratory. Regional patient data repositories or exchanges can serve many purposes: they can provide a source of reference information for past treatment, a basis for current patient data distribution in a geographically distributed health care environment, as well as a data depository for consultation services and workload distribution. In a normal medical practice, the various forms of data distribution complement each other. e-Referral and e-Discharge letters In 2005, the e-Referral letter and e-Discharge letter service was provided by 16 of the 21 hospital districts. Rapid progress was made over a period of a few years, because in 2003 the service was available in about half of the hospital districts. In primary care, 45% of the health care centres made use of the service in 2005, while the figure was 24% in 2003. Electronic and Televideo Consultations Electronic consultation was provided by 11 of the 21 hospital districts, and it was at testing or planning stages in the rest of the hospital districts, with the exception of one hospital district. About one third (36%) of the 174 health care centres purchased electronic consultations from hospitals. Another third reported planning or having even tested it. The health care centres purchasing electronic consultations seemed to be very active with its use: in 70% of the 49 health care centres which answered the question it was their principal mode for consultation. 72 In 2005, videoconferencing service was provided regularly by 10 of the 21 hospital districts, while five were in the testing or piloting phase. Among 179 health care centres that answered the questionnaire, 21 (12%) purchased video conferencing in order to consult a specialist at a hospital. In addition, ten health care centres had been planning or were testing it. Although specialist consultations by video conferencing were in 2005 provided by more hospitals than earlier, the number of health care centres purchasing the service had not increased from the 21 in 2003. Regional Data Exchange Systems According to the year 2005 survey, nine out of 21 hospital districts had a regional patient data repository in clinical use for the exchange of patient data and six districts were running pilot projects. There is no single closed network dedicated to healthcare in Finland; technically, the connections are provided through the use of commercial high-speed public data networks and secure tunnelling such as VPN over the public network. Teleradiology and Image Distribution through a Regional Archive The results from the surveys conducted in 2003 and 2005 on teleradiology and regional image distribution/archive services for secondary care (hospital districts) are presented in Table 6. The usage percentage is also given, if available. Since teleradiology services could be independent of local PACS or a regional archive, a combined look at image transfer services is given. The key information is that in 2005, 18 out of 21 (86%) hospital districts utilized some form of electronic distribution of radiological images. Table 6. Teleradiology and regional image distribution services in Finnish hospital districts in 2003 and 2005 (Paper VII, published by permission of IOS press). Measure 2003 2005 Teleradiology in production phase 13/21 16/21 Teleradiology in pilot phase 4/21 2/21 Teleradiology usage > 90% 2/21 5/21 Regional archive in production 3/21 10/21 Regional archive in pilot phase 0/21 3/21 Regional archive usage > 90% 0/21 3/21 Either teleradiology or regional archive in use 13/21 18/21 73 In the primary health care sector, the combined results show that 52 primary health care centres out of 179 (29%) had some method of teleradiological image delivery in production. This is a remarkable increase compared to 2003, when 15 primary health care centres were using teleradiological image delivery. Telelaboratory In 2005, regional distribution of laboratory results through a regional archive was utilized by 11 out of 21 Finnish hospital districts, one was at the pilot-project stage and three were planning to provide such a service in the near future. In addition, 16 out of 21 hospital districts utilized some other form of electronic transmission of laboratory results to the primary health care centres in their region. These other services were partially overlapping with the usage of the regional archives. The combined results showed that a total of 19 out of 21 (90%) hospital districts had some method for the electronic distribution of laboratory results in 2005. This figure has nearly doubled from 2003, when 10 hospital districts used telelaboratory services. In the primary health care sector, 48 health care centres out of 179 (27%) reported in 2005 that they received daily laboratory results electronically through a regional database, while 71 informed that they were either at testing or planning stage. Information Exchange between Health Care Organizations and Patients Twenty hospital districts and 79% of all the health care centres maintained their own websites. The hospital district and half of the health care centres who did not have an individual website were running pilot projects or were planning the implementation of their own website. Information exchange with patients by SMS messaging was used by one and tested by three of the hospital districts. It was used by nine (5%) health care centres and tested by 11 (6%) of the 179 health care centres. With private providers information exchange with patients by SMS was used by three (11%) and tested by five (18%) of the 28 private providers. Only one health care centre used secure e-mail in information networks, while 15% used conventional e-mail. Remote browsing of EPR by the patient was not in use anywhere, but was planned to be implemented by two hospital districts, three health care centres, and one private care provider. 74 6 Discussion This work was conducted during a time period when teleradiology technology advanced in great leaps. Therefore, the discussion topics touch a wide variety of implementation issues, both technical and clinical. The feasibility of commercial off-the-shelf technology for medical purposes is related to the clinically acceptable image quality level, but also to investment costs at that particular time. Radiological work practices have changed considerably because of teleradiology, but also because of the digital radiology environment in general. The Internet and mobile networks continuously give rise to questions of adequate transmission speed and required compression ratio of images. Likewise, portable terminals and even smaller smartphones challenge our attitude towards display characteristics acceptable for different purposes. Security, standardization and integration issues are crucial for all the implementations. Because radiology as a medical speciality is serving other disciplines, teleradiology should be investigated in relation to regulations and other digital and remote medical services in the eHealth domain. The keywords of the most common challenges are presented in table 7. Table 7. Common challenges when building a teleradiology service presented as keywords. Hardware and software related Communications related Teleradiology service related Cost of the equipment Availability of connections Work practice issues Quality of the images Speed of the network Medical quality of the service Usability of the equipment Quality of Service Liability, other legal aspects Image compression rate Cost of the network Patient´s consent Display characteristics Security of connections Reimbursement Integration and interoperability Mobility of the terminal devices Language issues It is interesting to note that these issues will come up again and again as teleradiology is finding new frontiers from point-to-point to an international and mobile setting. Because of the time span and different technologies used, each of the problems investigated in this study is first discussed separately in relation to the work where it first appeared, after which common findings and development trends are discussed. 75 6.1 Low-end technology teleradiology In the case of high-standard film scanning teleradiology systems with resolution equal to the original films and full 12-bit (4096 steps) grey scale with instant image transmission, the price for a single-site unit at the time of this study could be as high as USD 196,000 and a central referral unit may cost as much as USD 343,000. Network services make for additional costs if permanent connections are used (Batnitzky et al. 1990, Dwyer et al. 1991, Templeton et al. 1991). The cost of the equipment in this study was USD 20,000 for the transmitter and USD 15,000 for the receiver unit. This cost level made film-based teleradiology feasible for health centres and smaller hospitals. The kind of dial-up ISDN lines used in this study are available in all parts of Finland. Their fixed monthly costs are only USD 20 to 100, depending on the amount of simultaneous channels used, and the connections are paid by use, the price being that of normal long-distance calls. In their report on the use of low-cost, low-end digital teleradiology Lear et al. (Lear et al. 1989) built a system based on a prototype high-resolution CCD camera, a personal computer and an ISDN link. Their equipment and software were however proprietary and not commercially available. Dohrmann et al. (Dohrmann 1991) used personal computers to establish a low-cost teleradiology link in Australia. They used a video camera and a frame grabber to capture the images from printed films, however. A slow-speed (2,400 bit/s) ordinary telephone line modem was used to transfer images. This solution limits maximum resolution and transfer speed and is mostly suited for CT images. Yamamoto et al. (Yamamoto et al. 1992) built their low-cost teleradiology equipment out of personal computers and handheld scanners. Their scanning area was limited to 6.1 × 7.0 centimetres, and an ordinary telephone line was used for transmission. The on-line sending time for a 10:1 compressed chest image (size before compression less than 4 megabytes) in this study was at the same 30 sec level as reported by Goldberg et al. (Goldberg et al. 1993) for an uncompressed 7 megabyte image over a fast 1.544 Mbit/s network. If efficient compression can be used, even ISDN networks can serve effectively for this purpose. Only a few studies, however, discuss the total procedure time for teleradiology (Carey et al. 1989, Lear et al. 1989, Slovis & Guzzardo-Dobson 1991). In the present study, when efficient compression was used, film digitization together with other preparations took up more of the total procedure time than the actual sending. 76 The transmitted images were scored as technically acceptable with 8 exceptions (3%). In seven cases the inadequacy was due to improper scanner settings or insufficient grey scale dynamics. Because the scanner in the present study limited the image grey scale to 8 bits (256 levels), parameters such as brightness and contrast were critical for good image quality. If the original images present excessive contrast, details in both the dark and the light areas may be lost. This means that such images should be sent in duplicate for the dark and light areas. Another solution is an image scanning program which can decrease image contrast by making the highlights darker and the shadows lighter (thus modifying the response curve in the digitalization process). This method was used in phase three (from the regional hospital). True 12-bit (4096 levels) scanners that are not so operator-dependent are three to four times more expensive than the unit in this study. Image spatial resolution was sufficient, except in mammograms. The maximum scanning density of the tested system is 300 DPI, which equals 11.8 pixels per mm. This exceeded the scanning density of the Fuji AC-1 CR image plate system, for example. In practice, in the present study the scanning density was finally set to the same level as in the CR image plate system for chest (129 DPI = 5.1 pixels/mm) and bone (165 DPI = 6.5 pixels/mm) images. If the maximum scanning density of the tested scanner had been used all the time, the image sizes would have been too large for the equipment in this study. A 35 × 35 cm chest x-ray with 300 DPI (11.8 pixels/mm) resolution would require more than 17 megabytes and could not be loaded into the 8-megabyte memory of the computer. There were no significant differences in the diagnoses of technically acceptable chest or bone images. In two CT images, a small pons infarction and a slight thickening of the nervus opticus were missed in the original screen reading. In retrospect, however, these lesions were detected. It seems that the errors were due to the reader’s unfamiliarity with the computer screen. The opportunity of film printing might have been an advantage, but on the other hand, when picture archiving and communication systems are taken into use, computer workstations become a common workplace for radiologists. In mammograms, the error frequency was markedly greater, 10%, suggesting that the scanning resolution used in this study was not sufficient for transmission of the most detailed films. The results of this study, with their total 98% agreement on the technically acceptable images, correspond well to the results by Carey et al. (Carey et al. 1989), who sent 489 laser-digitized conventional and US examinations with 98% 77 agreement between the original and transmitted images. It is notable that they still had 12 patients in whom disease was not correctly identified even though the original films were digitized with 12-bit grey scale. Slovis et al. (Slovis & Guzzardo-Dobson 1991) also had an agreement of 98% in their work with 4,200 conventional paediatric images. Goldberg et al. (Goldberg et al. 1993) had discrepant interpretations in 18 (2.6%) of 685 cases while using a 12-bit scanner. When standard computer programs are adapted for clinical work, some limitations are encountered. The programs may not have been specifically designed for medical work and require many adjustments. In the first two phases of this study, the participants had to perform all the preparatory steps and make the changes from the viewing program to the transmission program. This was the reason why they let a computer operator send the images from the health care centre. This caused some quality problems because the computer operator was not accustomed to making decisions on the technical quality of scanned images. The easier graphical interface, the “shell” used in phase three, resulted in more convenient image transmission. It eliminated complicated steps and required less training. It made it possible for x-ray technicians to transmit the images as part of their normal daily routine without being too much concerned with the technology. This is important, because x-ray technicians cannot spend too much time on teleradiology efforts. The x-ray technicians were also able to maintain a constant image quality. This is important for the cost-effectiveness of the system. If a teleradiology system is based on standard low-end personal computers, one has to be aware of the limitations. Depending on the scanner, the grey scale dynamics and scanning density might not be sufficient for the most detailed images, such as mammograms. In addition, the smaller display unit necessitates the use of image zooming. If these problems could be resolved, the same central unit could be used for office automation, Medline literature searches and scientific image processing. Radiologists who are familiar with computers do not need to learn a new user interface, but can utilize the skills they already have. Images are also transferable into teaching programs and archives, for example. Our investigation suggested that good image quality can be restored in the transmission of images with the standard personal computers tested and a commonly available commercial software configuration. This could make it possible to build low-end technology teleradiology links for special purposes at a lower cost compared to the commercial high-end systems. However, this study was not specifically focused on making a detailed cost analysis of teleradiology. One should bear in mind that the fixed costs related to equipment are a function 78 of the equipment existing on the market at that time and all the economic conclusions are therefore related to the specific equipment and specific teleradiology setting. According to the experience in this study, the scanner is the most critical unit and its 8-bit grey scale should be extended. In this system, the total sending procedure still took a relatively long time, and the manoeuvrability of the system was thus another important limitation. Concerning clinical application areas, the system investigated here was not suitable for mammographic image transmission. 6.2 Work practise analysis New technologies necessarily provoke changes that go beyond the technical system itself. The work practice-oriented study of everyday clinical teleradiology work clearly illustrates that this is also the case with teleradiological systems, and it should be taken into account in designing and implementing them in practice. The study reveals important aspects that have previously gone unnoticed or been invisible (Plowman et al. 1995). For example, the articulation work done to ensure the smooth flow of the teleradiology service made up a major part of the work process both in terms of the amount of activities and importance for the whole process. Several aspects of cooperative work between the two geographically separate radiology communities were also identified. Moreover, articulation work and issues dealing with cooperation were neglected or underestimated in the system design and organization of teleradiology work practice. From radiologists’ point of view, it is important that the system does not interfere with the accustomed script of interpretation and reporting by causing disruptions and breakdowns. On the other hand, it has to be kept in mind that established work practices have been historically formed in close relation with associated technologies. Therefore, former working habits cannot always be kept in place as a standard way of working but should be critically evaluated. The work practice-oriented approach also proved to be useful in detecting those areas of work that are essential for the primary goal, i.e. accurate interpretation of images. The system should support the essential parts and typical tasks of work. Image comparison is an elementary component in radiologists’ work. Computer screen-based image interpretation technologies cannot supply as much display space as the radiologists are used to having with traditional light screens. However, the need for the system to allow for easy juxtaposition of images 79 remains. The lack of display space has to be compensated in another ways by the digitized images and computing power. Comparison could be made easier e.g. by always displaying chest studies on the screen in the same order. New display techniques, like stack or cine mode, could be used more often while reading a series of CT images. There is evidence in literature, that his will enhance the reading speed (Mathie & Strickland 1997, Kim et al. 2005b). Another aspect contributing to the ease of comparison of images is the image articulation work that is done as much as possible prior to the actual sessions. It remains to be seen to what extent this can be supported automatically by the system, as there seems to be a great deal of tacit knowledge involved in the articulation work carried out by the assistants. Unnecessary image arrangement tasks during the interpretation session should be avoided because they are time-consuming and can weaken the radiologist’s concentration on the image reading process itself. The initial image display should give an overview of the kind of images there are to be interpreted. A structured way of showing the time order of images is also useful. The interface should be simple, with only relevant buttons and menus, so as not to hinder the interpretation process. In the system studied here, the most often used image processing tools were adjustment of brightness and contrast as well as image zooming. Even though these features need training and take some time, they offer advantages by revealing diagnostic information that is otherwise not as easily noticeable. The investigated system had brightness/contrast control in the mouse, and some radiologists used it in an interactive way even while examining various parts within one image. More complicated image processing features seem not be needed in the busy routine praxis. Improvements to system usability were identified during the study. Some specific design guidelines for the development of the system were provided. However, this remains an area where more empirical work is needed, as the introduction of new technologies is always a process of learning and adaptation as well as innovation. The workshops helped a new work community emerge around the new system, as the simplest method is to let people from both ends meet each other in person! The personnel were able to discuss and compare their local procedures, and see for themselves how the work was carried out at the other end. This helped in building a shared understanding of the cooperative work process and in agreeing on common procedures – which is often a complicated issue with distributed systems. It is important to involve not only final end-users of the 80 system but also representatives of all of the tasks that will be changed by the system in the design process. This kind of approach is thus recommended when implementing new telemedical services. New technologies may introduce tasks that are altogether new; in this case such tasks included the scanning and sending of images in Kuusamo Primary Health Care Centre. It is important to adapt the new or modified tasks to local conditions and other practices that will exist in parallel when the system is used. Furthermore, it is important to identify what kinds of skills and knowledge are needed in the new tasks: not only training in skills for the new application programs or one’s own tasks is needed, but also knowledge of the tasks of others and one’s own role in the system as a whole to help in adjusting to team work. In the system studied, texts (requests and reports) and images were linked together. This approach is different from many teleradiology systems, where only images are transferred electronically while texts are sent via telefax (Goldberg et al. 1993, Davis 1997, Poca et al. 2004). The approach used in the studied system makes it easier to combine patient information and refer back to earlier cases. However – from the point of view of clinical praxis – an even more extended connection to patient records is preferred. Image comparison would be easier if earlier images could be acquired from a digital archive whenever needed. A hospital should maintain a common regional archive together with primary health care centres connected to it. This would increase the quality of care by making the diagnoses more accurate and also decrease unnecessary retakes as the patient moves from one institute to another. In addition, the report texts need to be integrated to the EPR. During the study, the primary health care centre printed out the reports in paper form. The goal is that physicians can automatically receive reports directly into their terminals. At that point, teleradiology would finally support medical work flow as part of vital patient information. 6.3 Internet teleradiology The value of this early study was that it showed the potential of the Internet, because only a few cases of international teleradiology had been reported earlier. This work also highlighted the common issues that are constantly relevant in connection with this type of distant communication. The technical quality of digitized and transmitted images was found to be adequate for diagnosis. A common image file format and common transmission protocols available for Internet connections made it possible to establish links 81 between different institutions in different countries. Standard data format developed for medical images (DICOM) will make this even easier in the future. The value of the Internet is its wide availability – connections are made using existing networks. This can help in sharing medical expertise in various part of the world in a way not experienced before and open new possibilities especially to the developing countries (Wootton et al. 2005). The openness of the Internet can on the other hand cause problems with data safety. Encryption of patient identification data is therefore needed in clinical use. For certain purposes even anonymized data can be used, where all patient ID information has been removed. The Swinfen Charitable Trust has successfully used the anonymization principle in their charity low-level telemedicine e-mail consultations for the third world since 1999 (Vassallo et al. 2001). In any case, hospital networks must also be protected by firewall systems against intruders from outside. Because of different transfer speeds between links, sending times can vary a lot. The speed of the fastest connection in the work discussed here was 2 Mbits/s and that of the slowest 9.6 kbit/s. To increase the throughput within such a slow channel, efficient data compression is necessary. Even though data transmission speeds have steadily increased within the Internet as well, and in 2010 Funet member organizations connect to the new routers with either 1 Gbit/s or 10 Gbit/s connections (CSC 2010), no QoS can be guaranteed. This is however not a major concern in the “store and forward” type of practice used in teleradiology. Commercial dedicated wide area networks have promised to fulfil the requirements of speed and data safety. Their problem is, however, that international connections are either not widely available or that they are still expensive. There are only few examples of dedicated health care networks that cross national borders, one of them being the Scandinavian health network (Duedal Pedersen 2004), which connects Medicom in Denmark, Carelink in Sweden and the Norwegian Healthnet. Therefore, secure connections over the Internet and compliance with the DICOM standard in file format and transmission have been used for offshore teleradiology services, e.g. examinations made in the US are read in reading centres in Australia and Switzerland (Bradley 2004, Bradley 2008). The effectiveness of a remote consultation system is not only a matter of time and technical image quality; informal additional information also needs to be negotiated during the consultation sessions. Special consultation software with multimodality capability (video, speech, cursor synchronization) is an advantage (Ligier et al. 1993). 82 In this study with a few real consultation cases a secondary sub-specialist opinion between university hospitals was considered useful, because the way medicine is performed was similar between those participated institutions. For small scale secondary opinion consultations between experts the language issue was not a problem, as medical English was used here. In the case of larger-scale consultations between the clinical physicians and radiologists the language issue becomes more relevant. E.g. in the European Baltic eHealth project the language issue was recognized as the biggest barrier to overcome in cross-border teleradiology (Ross et al. 2010). Therefore, they created a structured reporting tool to overcome this barrier and used that in consultations between the Baltic states and Denmark. Also multilingual transcriptionists were used. They also verified the medical quality of the reports before entering to production phase. Today, the success of cross-border teleradiology seems to be more dependent of organizational and quality issues than on pure technology. Also legal issues need more clarification (Stanberry 2006, Pattynama 2010). 6.4 Mobile PC teleradiology The benefits of teleradiology have been especially apparent in neurological emergencies (Lee et al. 1990, Spencer et al. 1991, Gray et al. 1997, Hazebroucq & Fery-Lemonnier 2004, Ashkenazi et al. 2007, Kreutzer et al. 2008,). Fixed connections like computer networks or telephone lines have mostly been used (Batnitzky et al. 1990, Dwyer et al. 1991, Stewart et al. 1992). This limits the mobility of a consulting senior to a certain place. Yamamoto published the idea of using wireless communication and portable computers for radiological consultation (Yamamoto 1995). However, analogue wireless communication involved transmission errors and the effective speed was slow. The digital GSM technology used in this study gives reliable connections and tolerable transmission speed with compressed images. In digital form the data remain unchanged. The relatively slow speed of the GSM network requires special techniques to keep total connection time on an acceptable level. To reduce the required bandwidth, we used JPEG compression for the images. In various studies it has been shown that a slight compression with a compression ratio from 1:10 to 1:15 does not affect diagnostic quality of radiological images (Lo & Huang 1986, Ishigaki et al. 1990, MacMahon et al. 1991). The system in this study was constructed using existing networking and image processing standards. Image 83 compression is also available for the DICOM format. Mobile telephone networks are becoming faster and thus more suitable for image consultations. The resolution and contrast of the active matrix display system of the portable computer used in the present study was sufficient for CT work. In particular, the relatively high luminance compared to older passive matrix screens was found to be suitable for CT reading. The active matrix display itself was capable of showing 256 colours with adjustable brightness. The screen in this investigation was better than in an earlier study, where a 16 grey-shade display was not found good enough to display X-ray images adequately (Yamamoto 1995). Modern laptop computers seem to be appropriate for clinical work. In this investigation, CT-images were interpreted with high reliability. One of the limitations in the system studied was display size. In one case the neuroradiologist would have liked to review the images side by side. An alternative is to use cine mode or a zoom tool of the image viewing software. Bigger flat screens have become available for portable computers, too. One of the reasons why experts considered some images only adequate for preliminary diagnosis was their inability to change the grey scale window and level. This is not related to the display technology, but to the file format used. In image file conversion, the original pixel depth is compressed from 16 to 8 bits. This means that in the studied system the images had to be windowed by the sender in order to optimize their visualization. This causes problems, if, for example, bone structure has to be interpreted and only soft tissue images have been sent. If DICOM format images would have been used, the receiver could have adjusted the images. In other cases the neuroradiologist was only able to give a preliminary diagnosis, because previous images would have been needed for comparison. With electronic archiving, these will be readily available on-line. This suggests that any teleradiology system should have a link with hospital PACS. In the study series, the two slight misdiagnoses caused no change in treatment. The first one appeared in the very first case in the study series and was due to uncertainty about the capabilities of the computer. This emphasizes the importance of individual experience with any new computer tool. The neuroradiologist was familiar with PC teleradiology, but had used a tabletop computer with an ordinary cathode ray tube (CRT) display. More aggressive use of brightness and contrast adjustment would have helped in making a correct diagnosis. The second case was difficult to decide even from the original image. 84 When widely-available network tools are used, data security is an important issue. The system studied here used a private dial-in server with user authentication requiring a password. No public computer networks, such as the Internet, were used. In the GSM network itself, the data are transmitted in encrypted digital form, and no additional encryption is thus needed. The usefulness of communication technology enables more efficient use of expert consultation. Patient care is more efficient and economical when a final diagnosis can be made at an early stage. This is very important in neurosurgical emergencies (Richter & Braun 1993). Consultants can react more promptly if they do not have to travel to a hospital to give their opinion. When hospital-wide image networks are built, remote links should also be considered. With mobile communication, experts can be reached if required, wherever they are – even in another country. This type of sub-specialist expertise could be useful in circumstances where the health care delivery system is similar, e.g. in the sparsely populated northern areas in the Nordic countries With modern technology, it is now possible for a radiologist to carry a toolbox for CT scan interpretation in a briefcase. The main limitation in wireless communication is the data transfer speed, but the situation is steadily improving. The presented initial study confirmed the feasibility of mobile PC teleradiology. 6.5 Smartphone teleradiology 6.5.1 Feasibility study Clinical results No studies were found in the literature on pocket-sized teleradiology terminals using intelligent phones or PDA-devices with combined digital phone capabilities in one device (a smartphone) before this investigation was accomplished. Wireless transfer of radiology images to a portable computer has been reported with analogue and digital technique (Yamamoto 1995). He also published at the same time a small series of 5 transmitted images to a pocket computer with a separate wireless modem connection (Yamamoto 2000). Image reading on a smartphone-type PDA-device was capable of serving as a preliminary reading station: no significant findings went unnoticed and a senior opinion could be given. A possibility to window and level the images might have 85 helped in detecting missed secondary findings. Actual reports showed a good agreement with reference film reading; only 3 cases had minor differences, although the neuroradiologist originally scored 20 cases in the category “preliminary report”. On the other hand, smaller findings can also go unnoticed in the case of film reading, as happened during this study for a small pons infarction, which was at first attempt only detected on the PDA terminal by the neuroradiologist. Using a PDA terminal was found suitable for reading the most common emergency CT findings like haemorrhages and brain infarctions, and it gives added value by enabling clinical consultations (by a neurosurgeon or neurologist) based on images. Technical limitations and expectations Although the PDA devices are small enough to be carried in the pocket, their networking and display technology must improve before they are more widely accepted as teleradiology terminals. Unlike many other telemedicine solutions, the cost is not a major barrier. Because off-the-shelf technology can be used, the cost of the terminal devices remains low. With the data call costs at the time of the study in Finland (0.20 Euros/min) the connection costs of a 30-minute session are also acceptable (6 Euros). Currently, there are several different vendors providing PDA devices on the market. Until recently, communication with these devices was complicated. The Nokia Communicator 9000 was the first device to successfully combine a PDA with a GSM-phone and make connectivity easy to use for everyone. The limited grey-shade resolution of the investigated PDA device was probably the limitation behind the difficulties to notice very small changes. Another limitation was the small screen pixel amount (even less than VGA). Both of these limitations can be partly solved by display software by providing tools for windowing and levelling the image as well as functions to zoom and scroll the image. Flat panel display technology itself is also evolving fast. Third-generation mobile terminals and phones will be capable of showing more grey-shades and pixels and even sharp colour images on a small display. If the power consumption of these new displays can be set to a reasonable level, image quality in a portable terminal will no longer be an issue. It is also possible to design the telemedicine service so that several optional devices can access the system. Some of the portable devices can also be connected to larger external display devices. 86 One might also ask how small displays could suffice in portable devices when at the same time a great deal of attention is given to the requirements and quality control of image reading displays in the radiology department and clinical wards. The answer may partially be related to the main purpose of these devices. For clinical decision-making, when there is no other option to get access to images, these devices provide a helping hand. In addition, many findings especially on cross-sectional CT and MRI images that are relevant to the clinical question at hand can be visualized with lower resolution requirements, if the tools are carefully used. One of the biggest restrictions of communicative PDA-technology in this study was the available bandwidth. The standard GSM data speed is 9.6 kbit/s, which restricts considerably the amount of data that can be transferred. The situation where multiple images in a CT scan series can be sent on average in 21 minutes is just within acceptable limits. However, one uncompressed chest X-ray image is 10MB in size, for example, and it would take several hours to transfer it over the GSM-network. In the work of Yamamoto and Williams (2000) viewing times from turning on the pocket computer to viewing the entire image ranged from 4 to 6 minutes, but only single images were sent, not series. Reliability in obtaining a wireless Internet connection was good, but not perfect. In the studied case with a dedicated device, reliability in connections was not considered a problem. Faster GSM network enables speeds up to 14.4 and 57.6 kbit/s. The first 14.4 kbit/s commercial service started in 1999. These higher speeds are achieved by high-speed circuit-switched data (HSCSD) technology by using multiple time slots for the data connection. Other approaches to speed up mobile data transmission in wide area networks are GPRS and EDGE. Finally, 3G (UMTS) technology with wireless data rates of up to 2 Mbit/s became commercially available in the mid-2000s, its latest high-speed downlink packet access (HSDPA) developments giving in 2010 a theoretical mobile data speed up to 14.4 Mbit/s depending on the capabilities of the network and terminal devices. There are also other global possibilities for wireless communication, such as mobile satellite services (Lamminen 1999). Their restriction compared to cellular networks is higher price and unavailability of integrated PDA devices. In summary, the spectrum of wireless communication for medical purposed covers radio pagers, radios, cellular telephones, satellite communication and personal communication services (Yoho 1994). 87 Rapid development in network technology means that mobile systems will come close to LANs in performance and that in the years to come, transmission times will no longer be a barrier, at least not in the vicinity of urban settlements. The latest development in mobile network family, fourth-generation mobile networks (4G) like LTE (long time evolution) wireless networks can theoretically provide downward links to the terminal with speeds of 100 Mbits/s, which is in the magnitude of fixed wired networks. The world premier for these networks took place in December 2009 in Stockholm and Oslo (Wikipedia 2010). For medical purposes, network security is an issue. The digital GSM system is inherently more secure than analogue systems due to its use of speech coding, digital modulation and TDMA (time division multiple access) channel access. The additional security mechanisms (such as authentication, encryption and temporary identification numbers) specified in the GSM standard make it one of the most secure telecommunications systems that are publicly available. Eavesdropping a GSM transmission is currently considered only a minor risk. Usage areas In emergency cases the ability to obtain advice for interpreting radiographs from a non-resident senior doctor seems to be very beneficial. It may be especially important in cases where decision on treatment options must be made in just a few hours. At present, PDA technology may not always be good enough for detailed diagnosis but it may be of great help in many clinical consultations where the radiological diagnosis is evident and the question concerns treatment actions. The majority of users may thus eventually be non-radiologists. The method was found suitable for clinical diagnosis of brain attacks. Intelligent phones, which can be carried in the pocket, make it possible to reach radiology consultants regardless of place. In the future, clinical consultations of e.g. neurosurgeons and neurologists will become easier, thus facilitating earlier treatment in the case of serious diseases like intracranial haemorrhages. Timely clinical specialist consultation has become more important not only in neurosurgical emergencies but also in the treatment of brain infarctions, where, according to the general guidelines, thrombolytic treatment should be started within three hours after the onset of symptoms (Käypä hoito 2006). It has been estimated that the main cause of delay in starting stroke treatment with thrombolytic drugs is the time required to perform the CT-scan. Part of the delay 88 may come from the delay of interpreting the examination. In future, the PDA terminal may speed up this process by providing direct access to the right expert. 6.5.2 Clinical study Network issues In the latter PROMODAS part of the clinical study, the faster download times between 5 to 10 minutes were achieved for a full series of CT or MRI scans using GPRS network technology. Even though mobile teleradiology in this manner is already feasible for special purposes and can be used clinically, standard mobile telecom networks still provide rather slow data transfer speeds. As mentioned earlier, the 3G networks (UMTS) and its 4G (LTE) successors are breaking this barrier by bringing real broadband to mobile devices. Another interesting wireless broadband technology is WLAN (Pagani et al. 2003), but its previously discussed geographical limitation of being available only in so-called service points, “hot spots”, means that it is usable only as an alternative or supplementary communication network for health professionals. Terminal device and software development Handheld devices are becoming more and more powerful tools. The computational capabilities are reaching the levels needed for many medical imaging requirements. The devices used in the PROMODAS project were already acceptable in terms of computational power. The image manipulation functions as well as rapid change of images were considered good enough for consultation purposes. However, as we are still in a phase of rapid development in mobile technology, the lifetime of mobile devices is rather limited. In practice, more than 3-year-old devices can be considered outdated. Because of the advances in network technology, but more importantly, because of the rapid progress of terminal devices and their operation systems, the development costs of consultation systems described in this study may rise compared to the expected lifetime of the applications. A new generation of devices every two to four years means challenges to both the technology provider and the user site. This problem can be alleviated by careful system design. For example, the DICOM standard can be considered a rather stable technology and it 89 should be used where possible. Similarly, general Internet-technology may be one solution: platform-independent applications may release us from platform dependence. The standardization of information transfer, contents and formats is needed for data exchange. At the time of this study, the MDD was not yet implemented to telemedicine software. In the future, the directive will have an effect on the development cycle of software for mobile devices. The evaluation of their safety and clinical feasibility has to be conducted in a regulated manner. However, in this study clinical evaluation was performed when diagnostic accuracy was compared to the original data on full-size image reading terminals. Most recently, PDAs have continued to advance with faster processors and faster wireless network connections, increased storage capabilities and improved screen resolutions. The Apple iPhone (Cupertino, CA, USA) released in 2007 has a screen resolution of 480 × 320 pixels in colour with high luminance LED backlight; its newest version, iPhone4 released in 2010, has a screen resolution of 960 × 640 pixels with a contrast ratio of 800:1 and a 500 cd/m2 brightness. A tablet size variant called iPAD already has a 1024 × 768 pixel resolution. In a recent study the luminance and contrast response of iPhone/iPodTouch displays was in the level of high-end commercial laptop displays and even medical display quality control phantom performance was adequate (Kondoh & Okuda 2010). High-resolution displays have also become standard in the Android and Symbian family of mobile phones. A new phenomenon in these operation systems is an easier way of making small applications (apps) to the new smartphones, so that there are already thousands of medical apps available. In addition, almost all PDAs now offer the ability to connect to standard WLAN networks, in addition to standard cellular networks. These portable devices will only continue to improve. Clinical experiences Also other research groups have shown the clinical usefulness of PDA devices in remote image consultation. Raman et al. (2004) have developed a PDA-based, DICOM-compatible teleradiology system that is capable of vendor-independent seamless integration with a PACS. Their system incorporates a Web interface and custom viewers for Palm OS and the Windows CE operating system, making it compatible with most PDA models available at that time. Their custom PDA image viewer software provides basic patient information and image management and viewing functions. The system supports compression and encryption and is 90 compatible with most wired and wireless communication methods. Although some hardware and software limitations still remain with respect to the development and implementation of the system, many of these limitations will be overcome in the near future thanks to ongoing development. Their experience showed that PDAs are capable of serving as a powerful platform for mobile image review and management. They expected that in the future, as high-speed wireless networks become ubiquitous, personal handheld devices will become an important platform for image review and reporting in the radiology department work flow and will enhance communication with referring clinicians. In their study, Yaghmai et al. (2004) found that PDAs are an effective way of transmitting 256 × 256 pixel resolution images through a PDA using the CDMA (code division multiple access) wireless network. These images were sent using standard e-mail clients, and the terminal used was a standard commercially available Treo-270 communicator PDA (palmOne, Milpitas, CA, USA). Instant transmission of radiological images may be important for making rapid clinical decisions about emergency patients. Kim et al. (2005a) examined an instant image transfer system based on a personal digital assistant (PDA) phone with a built-in camera. Images displayed on a picture archiving and communication systems (PACS) monitor can be captured directly by the camera in the PDA phone. Images can then be transmitted from an emergency centre to a remote physician via a wireless high-bandwidth network (CDMA 1 × EVDO, evolution data optimized). They reviewed radiological lesions in 10 normal and 10 abnormal cases produced by modalities such as CT, MR and DSA. The images were of 24-bit depth and 1144 × 880, 1120 × 840, 1024 × 768, 800 × 600, 640 × 480 and 320 × 240 pixels. Three neurosurgeons found that for satisfactory remote consultation a minimum size of 640 × 480 pixels was required for CT and MR images and 1024 × 768 pixels for angiography images. Although higher resolution produced higher clinical satisfaction, it also required more transmission time. At the limited bandwidth employed, higher resolutions could not be justified. The strength of this study was that it demonstrated that the mobility and portability resulting from the use of a personal digital assistant (PDA) in an emergency setting was beneficial, as it did not require the consultant to be available in a fixed location. However, the patient material was small compared to our clinical series. 91 Integration to medical record systems Teleradiology services can be utilized to their full potential only if they are integrated with EPR systems. Therefore the development of an initial mobile physician terminal was continued to clinical phase within the EU fourthframework projects MOMEDA and PROMODAS, both working with technology and process. One of the key elements in this project was to make technology more user-friendly, so that e.g. image transmission can take place in the background, without user interference. Thus it helped overcome the time delay spent on file managing found in the present study. The wireless terminal for a physician enables receiving of images, text and other components of multimedia patient record regardless of time and place. It is connected both to the radiology image network and to the comprehensive EPR system of Oulu University Hospital. The EPR is a multimedia record using WWW approach (Reponen et al. 2004), so additional information can be fetched with the common HTTP (hypertext transfer protocol) protocol. This means that the presented smartphone terminals gave full access to the medical record and not only to the radiological part of that. In this aspect the studied MOMEDA and PROMODAS terminals had an advantage compared to other studied terminals in literature. New models of work This development of personal wireless terminals may mean that some of the duty services can be centralized and that fewer physicians are needed during dutyhours. In the future, hospitals can directly consult e.g. neurosurgeons, neurologists or orthopaedic surgeons by sending images and referral letters to their portable devices. Ideally this will lead to re-engineering of current health care processes, freeing resources. Wireless workstations will also revolutionize the development of EPR in hospital. One of the biggest barriers of EPR development has been the cost caused by the distribution of enough workstations in the hospital to allow access to EPR from different work spots. Wireless workstations will give health care professionals liberty to move around and yet have fast access to the EPR using various wireless LAN-technologies and even switch between mobile data networks (3G) and WLAN with the same device. While the various barriers are lowering, the future telemedicine will be integrated tightly into the normal care processes and the technology that is needed will be in the routine toolbox of health care professionals. The patient is the final 92 winner when medical professionals can access information in the most practical way. 6.6 Integration of teleradiology as a part of eHealth services The present study is part of a continuing survey, which follows the implementation of the Finnish national eHealth roadmap. The ideas are based on the European eHealth action plan (2002), e.g. citizen-centred and patient-centred services, full continuum of care and the well-informed patient. This series of studies reveals data for administration and planning and even for international benchmarking. Teleradiology has been one of the first applications of telemedicine in Finland. According to a Finnish study, all the five Finnish university hospitals already had teleradiology services in 1994 (Reponen 1996). Regular service first started in the sparsely populated northern areas, but has since spread to all parts of the country. In the primary health care centres, the current trend is not to have a PACS of their own, but to combine efforts with a local regional hospital or with a larger hospital district. There are many innovative solutions available. For example in the most northern hospital district in Finland, all primary health care centres that produce x-rays are fully digitized and store their images at the central hospital. Those images can be accessed directly from the physician’s desktop through an EPR interface. In some areas small regional hospitals have a combined image archive and distribution system together with the primary health care centres. The boarder line between teleradiology and image distribution through a regional archive is gradually vanishing with certain services. In the current survey, all the methods used for a regional image transfer service were investigated. For a regional service, the basic assumption was that a hospital should have a local PACS installed. Then, the technical infrastructure behind the implementation of regional image distribution could differ. In some areas, images are viewed through a regional reference database. In other areas there is a dedicated common regional radiological database (“regional PACS”). A third solution is to view images through regional access to an EPR archive, which also contains images. The EPR is the most important tool for health professionals in acquiring and documenting information about patient care. During the year 2005 study in Finland, 95.6% of the primary health care centres and 95% of the hospital districts used EPR as their primary tool for patient data. Compared to the year 2003, there is very strong progress in secondary care both in terms of coverage in 93 various medical specialities and in the intensity of use. Because of the complexity of secondary care (hospital) medical records, the coverage aspect is an important indicator of EPR penetration there. Among primary health care centres, the proportion of health centres using EPR was already 93.6% in 2003. The most recent survey update in this series shows that 100% of Finnish hospital districts and practically 100% of health centres now have EPR in use (Hämäläinen et al. 2009). Knowing all this, one can say that the penetration of EPR is no longer a measure of ICT usage in Finland and that new indicators need to be sought. International comparison of eHealth status When comparing the Finnish health care ICT situation internationally, Finland belongs to the global leaders together with Denmark and Sweden, according to a recent report of the Information Technology and Innovation Foundation (Castro 2009). According to the report, all of these there countries are definitely ahead of the United States and most other countries in moving forward with their health information technology systems. This is because these three countries have nearly universal usage of electronic health records among primary care providers, high rates of adoption of electronic health records in hospitals, widespread use of health information technology applications, advanced telehealth programs and portals that provide access to online information. Similar high uptake of health ICT applications as in our study was also found in a study of ICT usage among general practitioners in Europe made by the research institute Empirica by assignment of the European Commission (Empirica 2008). They identified five eHealth frontrunner countries with the eHealth use indicator scoreboard: Denmark, the Netherlands, Finland, Sweden and the UK. Of these, Finland was the only country where all the primary care physicians used computers (and thus EPR) during their patient consultations. New indicators and challenges of eHealth implementation One new indicator is the exchange of information between health care institutions, mostly between primary and specialized health care. In Finland, regional exchange of information has been possible because of the high usage of EPR systems within institutions. Regional exchange of radiological and laboratory data has been one of the first applications, but today eReferrals and eDischarge letters directly from EPR to EPR are on the increase. All these contribute to the seamless 94 service systems of the patient. The popularity of regional data exchange has resulted in four different widely used data exchange systems. With the patient’s consent, the physician has instant access to previous patient data from other institutions through a secure connection. The exchange of electronic patient information between providers of health services necessitates the use of networks with high data security, which can be actualized through different kinds of intranet solutions or secure Internet connections. This data exchange is increasing rapidly in Finland. This is because digital data depositories in individual health care institutions are in active clinical use, and protected data connections enable the communication of electronic patient information. Electronic prescription has been mentioned as an important indicator of health care ICT maturity. In Finland, a national e-Prescribing pilot-project was launched in 2002. The pilot-project ran from 2004 to 2006. The system was tested in two hospital districts and a few health care centres. In 2007 a new law was issued (L 61/2007) where Finland opted for a system based on a national prescription database. In the pilot-project system, a doctor creates a prescription with a legacy system, signs it with a strong electronic signature, and sends the secured message to the national prescriptions database. The patient goes to a pharmacy where the pharmacist accesses the database with the pharmacy’s system. The pharmacist makes the required changes and marks the dispensing information on the electronic prescription, signs the markings with a personal smart card, and saves the markings to the prescription in the database and then delivers the medication to the patient. The clinical use of electronic prescription in Finland started in May 2010 in the city of Turku (STM 2010b). Once the backbone for electronic management of patient data in Finland is ready, the next major challenge is the construction and implementation of the National EPR Archive. The task is not a minor one. The stage is already set by the law on electronic management of patient information in social and health care (L 159/2007), which defines one secure national archive for permanent archiving of standardized medical data. Finland has taken a bold initiative and is now building a centralized, truly national solution. There are changes that need to be made, not only in technology but also in ways of conducting daily tasks in health care. According to the present law, the local systems should be connected to the archive by April 2011, but in practice a transition period of several years is needed. The greatest challenges may occur in communication between local and regional systems and the national archive. This might require major changes in the current 95 software being used in hospital districts and health care centres. The goal is increased access to electronic personal of health records, both by the physician and the patient. Update on national health care information architecture It has been a great challenge to implement the previously mentioned current legislation of electronic management of patient information (L 159/2007). That is why the Finnish government issued in autumn 2010 a new proposal for a modified act that would enable the national medical record archive to serve patient care better and make it easier for health care personal to have access to the required health information. It is also proposed that the implementation timetable should be adjusted. (HE 176/2010, STM 2010a.) According to the proposal, the medical record archive (eArchive) will be constructed stepwise in several phases. The contents will be further regulated by a decree of the Ministry of Social Welfare and Health. According to the current document, the first phase includes the so-called principal data, which contain medical narrative texts, nursing summary information, risk information, pharmaceuticals, laboratory results, radiological requests and reports as well as electronic referral and discharge letter contents. (STM 2009, KanTa 2010, STM 2010a.) Concerning medical imaging, it is important to note that initially only textual information (requests and reports) contained in the main patient record system will be archived. If the information is residing only in external systems (RIS), it is not obligatory to store it in the national archive. Medical images are not included in the first phase of archived material, but their turn will come in the next phase. As of autumn 2010 the enterprise architecture will be investigated and a proper solution will be suggested according to an expert consultation. The task will probably include discussion of whether images should also be stored at a single storage site, or whether regional storages should also have a role and what the relationships will be between them. At the same time, the processes concerned and therefore functional data architecture will be investigated. The extra time will give an opportunity to learn more about international experiences in this field. At the same time, the method of obtaining patient’s consent for delivery of electronic material from the national archive is made easier and more flexible. Patients can in the future provide access to their information to all professionals 96 involved in their care, even in different institutions. The authorization is valid until further notice, but patients can always cancel it. Of course, patients can prevent the delivery of information that is too sensitive (opt-out principle). While patients can use secure web access to view their most important health data (eView) they can in the future use the same tool to manage their consent (eConsent). (HE 176/2010.) The timetable will be adjusted according to current reality. According to the government proposal, all public health care organisations should have the first phase finished by 9/2014, and the private sector should follow by 9/2015. This means that there will be an extension of three to four and a half years compared to present regulations. (HE 176/2010.) Finally, more resources will be devoted to accomplishing the task. The changes include shifting operational steering of national health information system development and implementation from the Ministry to the National Institute for Health and Welfare. (HE 176/2010.) 6.7 General discussion – from teleradiology to eHealth and beyond During this study, technical platforms for teleradiology services were designed and built for a regional service setting, for international communication settings and for various mobile usage settings. Of these, no mobile smartphone combined teleradiology and EPR terminal was found in the medical literature before this study, and the international teleradiology connection using the Internet was also one of the early experiments in this area. Their evaluation consisted of the assessment of technical feasibility and diagnostic accuracy. Work practice issues with teleradiology were also studied in the early stage of this discipline, and finally, the distribution of teleradiology in the Finnish telemedicine and eHealth context was investigated in a nationwide comprehensive survey targeting health care institutions. One of the strengths of the study was that the concepts developed in this study were taken into clinical usage, or resulted in better system developments. The joint lessons learned from the regional low-end teleradiology study and the work practice study influenced the development of teleradiology scanners and workstation software. Finally, mobile smartphone teleradiology consultation has now been in clinical use at Oulu University hospital for 10 years. From a technical point of view, this study showed that common consumer technology in personal computers and communication networks is good enough 97 for medical purposes if performance aspects are discussed. Rosset et al. (2005) and Ratib et al. (2009) have made observations on the same topic. Of course, regulations and reliability requirements mean that medical device certification is needed for those implementations that are finally taken into permanent patient care. However, this general development has decreased the price of medical technology; especially powerful computers for medical image processing are now available at affordable prices. The value of standardization cannot be overestimated. During the time course of this study the DICOM standard was gradually becoming implemented in all areas of teleradiology, making data compatibility an easier task. However, in practical workflow issues more specific definitions are needed, and hopefully IHE profiles will be used more for teleradiology, too. In the Western Health Region in Norway, they have been able to connect numerous RIS/PACS solutions from different public and private institutions using IHE radiological cross enterprise document sharing definitions (Stoerkson & Aslaksen 2010). Another set of de facto standards that have simplified data communication have been the Internet protocols, which have made it easier to design data transfer tasks. In Swedish Västra Götaland region (VGR) they have been able to develop a joint teleworkspace which is integrated to the various RIS and PACS systems with networked tools and graphical user interfaces (Lundberg et al. 2010). According to the authors, the need for teleradiology as a service provided "by somebody" has disappeared in VGR; today it is a shared service embedded in the innovative radiology information infrastructure. In this investigation, data compression was used a lot after its usefulness was initially assured. Data compression is still necessary for teleradiology because of the narrow bandwidths used especially in mobile or home connections. Actually, successful use of data compression has been an enabling factor for new frontiers. Large datasets in cross-sectional imaging can also be one reason to use data compression. Fortunately, in literature it has been shown that diagnostic quality of the images is not compromised by moderate compression levels, not even lossy compression in the magnitude of 1:10 to 1:25. This naturally depends of image type and size and in Germany they have published recommendation to use a modest lossy compression in the magnitude of 1:5 to 1:15 depending on the target modality (Loose et al. 2009). The Internet provided a pathway and model for international teleradiology consultations. This investigation covered the most typical user case, i.e. the need of expert consultation for rare cases. It is worthwhile to notice that these 98 international consultations can be arranged in an easy and affordable manner. Because data security is an issue, it is important to note that today there are ways of setting up secure private connections over public networks. International teleconsultations, however, have certain aspects that go beyond the technical ones. The discussion of medical competence required and moreover, the discussion of whether radiological image interpretation is merely “image reading” or highquality medical specialist consultation is an ongoing debate. Recent advances in medical ICT usage have taken electronic patient records to the level where more tight integration between imaging systems and HIS can be predicted. Finland is one of the countries where EPRs have reached in practice 100% coverage and where picture archiving and communications systems are used comprehensively in health care. This study has revealed important new indicators, such as interorganizational data exchange, that can be used for benchmarking health care ICT services. It this setting regional image distribution and teleradiology have an increasing role as one of the eHealth services. Most of the technical challenges of teleradiology are solved little by little as the general digitalization of the imaging process and patient records progresses. Solutions are also gradually being found for a common digital working environment and eHealth-related workpractice issues in the most common teleradiology settings. However, work practices should be carefully adjusted in every new type of service setting. In its essence, teleradiology is a matter of conducting computer-assisted and remote co-operative work, either within the teleradiology work chain itself or in the longer medical work-up between patients, practicing physicians and teleradiology consultants. With increasing possibilities of implementing workload off-shoring and even international teleradiology, medical quality of the services and issues related to legal aspects, such as liability issues and patient’s consent will became more important. For international teleradiology, the issue of different languages and medical tradition remains a future challenge. The major development steps and driving forces for teleradiology have been discussed in table 8. 99 Table 8. The major development steps in teleradiology related to enabling factors in general ICT, medical ICT and the interests of various stakeholders. Decade Achievements Enabling factors Acting forces 1900 First ideas Telephone, telegram, radio Individuals 1920 First experiments Telegraphs, image facsimiles Individuals 1950 Clinical videoteleradiology Television, video cable Individuals, hospitals networks 1980 Commercial camera-on-stick Affordable video cameras, Individuals, hospitals, devices, early computer-based modems, first digital imaging research institutions, systems modalities, computer commercial enterprises workstations 1990 Commercial high-end systems, PC revolution, PACS and digital Reseach institutions, advent of PC-based teleradiology, imaging, standardization clinical operation, point-to-point- (DICOM), digital data networks, enterprises hospitals, commercial services, international experiments the Internet Teleradiology-PACS integration, Acceptance of DICOM Commercial enterprises, teleradiology service providers, standard, large scale PACS, health care providers, mobile systems development, fast data networks, mobile research institutions international pilots networking and computing, 2010 Practice of remote image reading Integration of PACS-RIS-HIS Governments and other and and distribution (“virtual hospital”), with IHE profiles, eHealth public stakeholders, promotion, ubiquitous health care providers, 2000 affordable data security beyond workload off-shoring, to-the-point mobile distribution, service quality computing, government interest commercial enterprises management Teleradiology as a concept has already moved on from its technical background. It can be used to denote a service within public health care, e.g. between primary care and secondary care hospitals, or it can be seen as an external service, where part of the work has been outsourced to external (private) providers. There are already services like mammography screening which mostly rely on teleradiology in image interpretation. One could say that teleradiology has become a fully-fledged tool in eHealth service delivery. Medical imaging results can be distributed, and so can image interpretation tasks. Ubiquitous imaging is already here. 100 7 1. 2. 3. 4. 5. 6. Summary and Conclusions This investigation suggested that a good image quality can be restored in the transmission of images with the low-end personal computer and a commonly available commercial software configuration tested. This could make it possible to build teleradiology links for special purposes at a lower cost compared to the commercial high-end systems. There were limitations in technical quality and manoeuvrability, however. Teleradiology between two institutions changed the image interpretation work processes and revealed hidden aspects of work practice that are not normally noticed. Especially articulation work that supports the key tasks is now mostly conducted at the receiving site. The receiving site is also heavily dependent on the decisions made at the sending site. The radiologists have to rely on less information than in the case of on-site image reading. Communication of relevant information is essential in both directions. This study showed that with careful planning the Internet can provide a feasible transmission channel for international teleradiology consultations between hospital networks. Even though the Internet is readily available, there are shortcomings in its speed and safety. Therefore, further investigation and development is needed, especially in the field of data safety. According to this study, a radiologist can carry his workstation for CT interpretation as a laptop computer in his briefcase. The main limitation was the speed of data transfer, but this situation will improve steadily with technological developments. Ubiquitous communication technology enables a more efficient use of expertise. However, the new tools require training in order to be utilized to full extent. Smartphones offer a new tool to gain access to patient information outside the hospital. Despite the small screen, the image quality in the terminals is sufficient for clinical decision-making, and even with low bandwidth the data transfer speed is tolerable. With the progress in technology, these devices seem to become more useful. Teleradiology has become an integrated part of eHealth and its distribution together with PACS distribution has steadily increased in Finland during the 2000s. From primary health care users’ point of view it seems to be irrelevant whether the image transfer services are called teleradiology or regional imaging study results delivery. 101 102 References Abbot KR & Sarin SK (1994) Experiences with Workflow Management: Issues for the Next Generation. Proceedings of the Conference on Computer Supported Cooperative Work. NY, ACM: 113–120. Ahonen P (2008) Funet Suomen tie internetiin. Helsinki, CSC - Tieteen tietotekniikan keskus Oy. Alaoui A, Collmann J, Nguyen D, Lindisch D, Nishant Subbiah RT, Green A, Johnson J & Mun SK (2003) Implementing a secure Teleradiology system using the internet. Int Congr Ser 1256: 803–808. Ali-Yrkkö J & Hermans R (2004) Nokia: A giant in the Finnish innovation system. In Schienstock G (ed) Embracing the Knowledge Economy - The Dynamic Transformation of the Finnish Innovation System. Cheltenham, UK, Edward Elgar Publishing Ltd: 106–127. American College of Radiology (ACR) (2002) ACR Standard for Teleradiology. URL: http://imaging.stryker.com/images/ACR_Standards-Teleradiology.pdf. Cited 2010/10/31. Andrus WS, Dreyfuss JR, Jaffer F & Bird KT (1975a) Interpretation of roentgenograms via interactive television. Radiology 116(1): 25–31. Andrus WS, Hunter CH & Bird KT (1975b) Remote interpretation of chest roentgenograms. Chest 67(4): 463–468. Anonymous (1907) Sending photographs by telegraph. The New York Times Part Three (February 24): SM7. Anomymous (1929) Sending dental X-rays by telegraph. Dent Radiogr Photogr 2(2): 16. Arenson RL, Andriole KP, Avrin DE & Gould RG (2000) Computers in imaging and health care: now and in the future. J Digit Imaging 13(4): 145–156. Ashkenazi I, Haspel J, Alfici R, Kessel B, Khashan T & Oren M (2007) Effect of teleradiology upon pattern of transfer of head injured patients from a rural general hospital to a neurosurgical referral centre. Emerg Med J 24(8): 550–552. Batnitzky S, Rosenthal SJ, Siegel EL, Wetzel LH, Murphey MD, Cox GG, McMillan JH, Templeton AW & Dwyer SJ,3rd (1990) Teleradiology: an assessment. Radiology 177(1): 11–17. Beckmann EC (2006) CT scanning the early days. Br J Radiol 79(937): 5–8. Benjamin M, Aradi Y & Shreiber R (2010) From shared data to sharing workflow: merging PACS and teleradiology. Eur J Radiol 73(1): 3–9. Bergh B, Pietsch M & Vogl TJ (2000) Internet technology for clinical applications in a digital radiography department. Radiologe 40(4): 371–387. Bergh B, Schlaefke A, Frankenbach R & Vogl TJ (2004) Remote access to a web-based image distribution system. Rofo 176(6): 896–901. Bergh B, Schlaefke A, Pietsch M, Garcia I & Vogl TJ (2003) Evaluation of a "no-cost" Internet technology-based system for teleradiology and co-operative work. Eur Radiol 13(2): 425–434. 103 Bergmo TS (1996) An economic analysis of teleradiology versus a visiting radiologist service. J Telemed Telecare 2(3): 136–142. Berry RF & Barry MH (1998) Evaluation of a personal-computer-based teleradiology system serving an isolated Canadian community. Can Assoc Radiol J 49(1): 7–11. Boland GW (1998) Teleradiology: another revolution in radiology? Clin Radiol 53(8): 547–553. Boland GW (2008) Teleradiology coming of age: winners and losers. AJR Am J Roentgenol 190(5): 1161–2. Bowers J, Button G & Sharrock W (1995) Workflow from Within and Without: Technology and Cooperative Work on the Print Industry Shopfloor. In Marmolin H, Sundblad Y & Schmidt K (eds) Proceedings of the Fourth European Conference on Computer-Supported Cooperative Work. Dordrecht, Kluwer: 51–66. Bradley WG (2004) Offshore teleradiology. J Am Coll Radiol 1(4): 244–248. Bradley WG,Jr (2008) Off-site teleradiology: the pros. Radiology 248(2): 337–341. Branstetter BF,4th (2007) Basics of imaging informatics: part 2. Radiology 244(1): 78–84. Calder LD, Maclean JR, Bayliss AP, Gilbert FJ & Grant AM (1999) The diagnostic performance of a PC-based teleradiology link. Clin Radiol 54(10): 659–664. Canadian Association of Radiologists (2008) CAR Standards for teleradiology URL: http://www.car.ca/Files/Standard_Teleradiology_EN.pdf. Cited 2010/10/31. Caramella D (1996) Teleradiology: state of the art in clinical environment. Eur J Radiol 22(3): 197–204. Carey LS, O'Connor BD, Bach DB, Hobbs BB, Hutton LC, Lefcoe MS, Lyons RO, Munro TG, Paterson RG & Rankin RN (1989) Digital teleradiology: Seaforth - London network. Can Assoc Radiol J 40(2): 71–74. Castro D (2009) Explaining International IT Application Leadership: Health IT. The Information Technology and Innovation Foundation Report September 2009. URL: http://www.itif.org/files/2009-leadership-healthit.pdf. Cited 2010/10/31. Cawthon MA, Goeringer F, Telepak RJ, Burton BS, Pupa SH, Willis CE & Hansen MF (1991) Preliminary assessment of computed tomography and satellite teleradiology from Operation Desert Storm. Invest Radiol 26(10): 854–857. Channin DS (2001) Integrating the Healthcare Enterprise: a primer. Part 2. Seven brides for seven brothers: the IHE integration profiles. Radiographics 21(5): 1343–1350. Commission Communication (COM/2002/0263 final) (2002) Communication from the Commission to the Council, the European Parliament, the Economic and Social Committee and the Committee of the Regions. eEurope 2005: Information society for all. An Action Plan. URL: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri= COM:2002:0263:FIN:EN:PDF. Cited 2010/10/31. Commission Communication (COM/2008/0689 final) (2008) Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions on telemedicine for the benefit of patients, healthcare systems and society. URL: http://eur-lex.europa.eu/LexUriServ/ LexUriServ.do?uri=COM:2008:0689:FIN:EN:HTML. Cited 2010/10/31. 104 Commission Proposal for a directive 2008/0142 (2008) Proposal for a Directive of the European Parliament and Council on the application of patients' rights in cross-border healthcare. URL: http://ec.europa.eu/health/ph_overview/co_operation/healthcare/ docs/COM_en.pdf. Cited 2010/10/31. Commission Recommendation C(2008)3282 (2008) Recommendation of 2 July 2008 on cross-border interoperability of electronic health record systems (notified under document number C(2008)3282). Official Journal L 190, 18/07/2008, 0037–0043. CSC – IT center for science (2010) Funet gets faster and even more reliable. News 04.03.2010. URL: http://www.csc.fi/english/csc/news/news/Funet-upgrade-en. Cited 2010/10/31. Curtis DJ, Gayler BW, Gitlin JN & Harrington MB (1983) Teleradiology: results of a field trial. Radiology 149(2): 415–418. Daucourt V, Petitjean ME, Chateil JF & Michel P (2005) Evaluation of the benefits for the patient arising from an inter-hospital teleradiology network in a French administrative area. J Telemed Telecare 11(4): 178–184. Daucourt V, Sicotte C, Pelletier-Fleury N, Petitjean ME, Chateil JF & Michel P (2006) Cost-minimization analysis of a wide-area teleradiology network in a French region. Int J Qual Health Care 18(4): 287–293. Davis MC (1997) Teleradiology in rural imaging centres. J Telemed Telecare 3(3): 146– 153. DeCorato DR, Kagetsu NJ & Ablow RC (1995) Off-hours interpretation of radiologic images of patients admitted to the emergency department: efficacy of teleradiology. AJR Am J Roentgenol 165(5): 1293–1296. Directive 2007/47/EC (2007) Directive 2007/47/EC of the European Parliament and of the Council of 5 September 2007 amending Council Directive 90/385/EEC on the approximation of the laws of the Member States relating to active implantable medical devices, Council Directive 93/42/EEC concerning medical devices and Directive 98/8/EC concerning the placing of biocidal products on the market. Official Journal L 247, 21/09/2007, 0021 - 0055. Directive 90/385/EEC (1990) Council Directive 90/385/EEC of 20 June 1990 on the approximation of the laws of the Member States relating to active implantable medical devices. Official Journal L 189, 20/07/1990, 0017 – 0036. Directive 93/42/EEC (1993) Council Directive 93/42/EEC of 14 June 1993 concerning medical devices. Official Journal L 169, 12/07/1993, 0001 - 0043. Directive 95/46/EC (1995) Directive 95/46/EC of the European Parliament and of the Council of 24 October 1995 on the protection of individuals with regard to the processing of personal data and on the free movement of such data. Official Journal L 281, 23/11/1995, 0031 – 0050. Dohrmann PJ (1991) Low-cost teleradiology for Australia. Aust N Z J Surg 61(2): 115– 117. Dripps JH, Boddy K & Venters G (1992) Telemedicine: Requirements, Standards and Applicability to Remote Care Scenarios in Europe. In van Goor JN & Christensen JP (eds) Advances in Medical Informatics. Amsterdam, IOS Press: 340–348. 105 Duedal Pedersen C (2004) E-health in the Scandinavian countries. Stud Health Technol Inform 106: 137–143. Dunne MG & Dunne SB (2004) Improved teleradiology reporting using fax software. AJR Am J Roentgenol 182(4): 871–874. Dwyer SJ,3rd, Templeton AW & Batnitzky S (1991) Teleradiology: costs of hardware and communications. AJR Am J Roentgenol 156(6): 1279–1282. Eklof H, Radecka E & Liss P (2007) Teleradiology Uppsala-Sydney for nighttime emergencies: preliminary experience. Acta Radiol 48(8): 851–853. Elisa Oyj (2004) Radiolinja´s history. http://www.elisa.com/on-elisa/corporate/ history/radiolinjas-history/. Cited 2010/10/31. Empirica (2008) Benchmarking ICT use among General Practitioners in Europe - Final Report by Empirica with assignment of European Commission, Information Society and Media Directorate General. URL: http://ec.europa.eu/information_society/ eeurope/i2010/docs/benchmarking/gp_survey_final_report.pdf. Cited 2010/10/31. Engelmann U, Schroeter A, Schwab M, Eisenmann U, Vetter M, Lorenz K, Quiles J, Wolf I, Evers H & Meinzer HP (1999) Borderless teleradiology with CHILI. J Med Internet Res 1(2): E8. Engelmann U, Schwab M, Schroter A, Rusu P & Meinzer HP (2002) Evaluation of CHILI teleradiology network 4 years after clinical implementation. Radiologe 42(2): 87–93. Engeström Y (1996) Developmental work research as educational research: looking ten years back and into the zone of proximal development. J Nord Educ Res 16(3): 131– 143. epSOS (2008) Welcome to epSOS – a European eHealth Project. (2008) URL: http://www.epsos.eu/. Cited 2010/10/31. Ericsson (2009) The evolution of EDGE - White Paper. URL: http://www.ericsson.com/ res/docs/whitepapers/evolution_to_edge.pdf. Cited 2010/10/31. European Society of Radiology (ESR) (2004, reissued 2006) ESR brochure-I. Teleradiology - white paper. URL: http://www.myesr.org/html/img/pool/ ESR_2006_I_Telerad_Web.pdf. Cited 2010/10/31. European Society of Radiology (ESR) (2006) ESR brochure-VII. Teleradiology in the European Union - white paper. URL: http://www.myesr.org/html/img/pool/ ESR_2006_VII_Telerad_Summary_Web.pdf. Cited 2010/10/31. European Society of Radiology (ESR) (2010) Expert Panel on Image Compression. URL: http://www.myesr.org/cms/website.php?id=/en/research/expert_panel_on_image_com pression.htm. Cited 2010/10/31. Gayler BW, Gitlin JN, Rappaport W, Skinner FL & Cerva J (1981) Teleradiology: an evaluation of a microcomputer-based system. Radiology 140(2): 355–360. Gitlin JN (1986) Teleradiology. Radiol Clin North Am 24(1): 55–68. Glinkowski W & Kornacki M (2002) Tecnological standards of teleradiology function in orthopedics and trauma care. Ortop Traumatol Rehabil 4(6): 725–730. Goh KY, Lam CK & Poon WS (1997a) The impact of teleradiology on the inter-hospital transfer of neurosurgical patients. Br J Neurosurg 11(1): 52–56. 106 Goh KY, Tsang KY & Poon WS (1997b) Does teleradiology improve inter-hospital management of head-injury? Can J Neurol Sci 24(3): 235–239. Goldberg MA, Rosenthal DI, Chew FS, Blickman JG, Miller SW & Mueller PR (1993) New high-resolution teleradiology system: prospective study of diagnostic accuracy in 685 transmitted clinical cases. Radiology 186(2): 429–434. Graschew G, Roelofs TA, Rakowsky S, Schlag PM, Heinzlreiter P, Kranzlmuller D & Volkert J (2006) New trends in the virtualization of hospitals--tools for global eHealth. Stud Health Technol Inform 121: 168–175. Gray WP, Somers J & Buckley TF (1997) Report of a national neurosurgical teleradiology system. J Telemed Telecare 3 Suppl 1: 36–37. GSM Association (2009) GSM_WorldPoster2009A. URL: http://www.gsmworld.com/ roaming/GSM_WorldPoster2009A.pdf. Cited 2010/10/31. GSM Association (2010) History ~ GSM World. http://www.gsmworld.com/aboutus/history.htm. Cited 2010/10/11. Halvorsen PA & Kristiansen IS (1996) Radiology services for remote communities: cost minimisation study of telemedicine. BMJ 312(7042): 1333–1336. Hämäläinen P, Reponen J & Winblad I (2007) eHealth of Finland: check point 2006. Reports 1/2007, National Research and Development Centre for Welfare and Health, Helsinki. Hämäläinen P, Reponen J & Winblad I (2009) eHealth of Finland: Check point 2008. Reports 1/2009, National Institute for Health and Welfare, Helsinki. Hanley JA & McNeil BJ (1982) The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 143(1): 29–36. Hanley JA & McNeil BJ (1983) A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 148(3): 839–843. Harno K (2004) UUMA. Regional eHealth services in the hospital district of Helsinki and Uusimaa (HUS). Stud Health Technol Inform 100: 101–108. Harno K, Gronhagen-Riska C, Pohjonen H, Kinnunen J & Kekomaki M (2002) Integrated regional services: are working process changes desirable and achievable? J Telemed Telecare 8 Suppl 3(6): 26–28. Hazebroucq V & Fery-Lemonnier E (2004) The value of teleradiology in the management of neuroradiologic emergencies. J Neuroradiol 31(4): 334–339. HE 176/2010 (2010) Hallituksen esitys Eduskunnalle laeiksi sosiaali- ja terveydenhuollon asiakastietojen sähköisestä käsittelystä annetun lain sekä eräiden muiden lakien muuttamisesta. URL: http://www.finlex.fi/fi/esitykset/he/2010/20100176. Cited 2010/10/02. Henri CJ, Rubin RK, Cox RD & Bret PM (1997) Design and implementation of World Wide Web-based tools for image management in computed tomography, magnetic resonance imaging, and ultrasonography. J Digit Imaging 10(3 Suppl 1): 77–79. Herscovici N, Christodoulou C, Kyriacou E, Pattichis MS, Pattichis CS, Panayides A & Pitsillides A (2007) m-health e-emergency systems: Current status and future directions. IEEE Antennas and Propagation Magazine 49(1): 216–231. 107 Huang HK (2010) PACS and Imaging Informatics: Basic Principles and Applications, Second Edition. Hobogen NJ USA, John Wiley and Sons. Huang HK, Mankovich NJ, Taira RK, Cho PS, Stewart BK, Ho BK, Chan KK & Ishimitsu Y (1988) Picture archiving and communication systems (PACS) for radiological images: state of the art. Crit Rev Diagn Imaging 28(4): 383–427. Hussein R, Engelmann U, Schroeter A & Meinzer HP (2004) Implementing a full-feature PACS solution in accordance with the IHE technical framework: the CHILI approach. Acad Radiol 11(4): 439–447. IHE (2010) IHE Profiles. URL: http://www.ihe.net/Profiles/. Cited 2010/10/31. IHE Europe (2010) Connectathon Results Browsing. URL: http://connectathon-results.iheeurope.net/. Cited 2010/10/31. Ishigaki T, Sakuma S, Ikeda M, Itoh Y, Suzuki M & Iwai S (1990) Clinical evaluation of irreversible image compression: analysis of chest imaging with computed radiography. Radiology 175(3): 739–743. Istepanian RH, Woodward B, Balos PA, Chen S & Luk B (1999) The comparative performance of mobile telemedical systems based on the IS-54 and GSM cellular telephone standards. J Telemed Telecare 5(2): 97–104. Istepanian RH, Woodward B, Gorilas E & Balos PA (1998) Design of mobile telemedicine systems using GSM and IS-54 cellular telephone standards. J Telemed Telecare 4(Suppl 1): 80–82. Jelaso DV, Southworth G & Purcell LH (1978) Telephone transmission of radiographic images. Radiology 127(1): 147–149. Johansen I & Breivik E (2004) Is teleradiology service in primary health care costeffective? Tidsskr Nor Laegeforen 124(19): 2490–2492. Jones J (2000) Britain's first "virtual hospital" gets go ahead. BMJ 320(7244): 1227A. Joutjärvi T, Suihkonen O, Airo V (2000) High Speed Circuit Switched Data (HSCSD). Seminaaritutkielma. Lappeenrannan teknillinen korkeakoulu, Tietotekniikan osasto, 22.2.2000. URL: http://www2.it.lut.fi/kurssit/99-00/010651000/seminaarit/ hscsd_final.doc. Cited 2010/10/31. Judén J & Kuokkanen A (2000) 1. ja 2. sukupolven matkaviestinverkot. Seminaaritutkielma. Lappeenrannan teknillinen korkeakoulu, Tietotekniikan osasto, 04.05.2000. URL: http://www2.it.lut.fi/kurssit/99-00/010651000/seminaarit/ 1ja2sukupmatkapuh.doc. Cited 2010/10/31. Jutras A (1959) Teleroentgen diagnosis by means of video-tape recording. Am J Roentgenol Radium Ther Nucl Med 82: 1099–1102. Kajiwara K (1992) JPEG compression for PACS. Comput Methods Programs Biomed 37(4): 343–351. Kalyanpur A, Neklesa VP, Pham DT, Forman HP, Stein ST & Brink JA (2004) Implementation of an international teleradiology staffing model. Radiology 232(2): 415–419. KanTa (2010) Kansallinen Terveysarkisto (KanTa) (2010) Ajankohtaista. URI: https://www.kanta.fi/web/fi/ajankohtaista. Cited 2010/10/02. 108 Karasti H (1997) Bridging the analysis of work practice and system redesign in cooperative workshops. Proceedings of DIS'97: Designing Interactive Systems: Processes, Practices, Methods, & Techniques 1997. NY, ACM Press: 185–195. Karasti H (2001) Increasing sensitivity towards everyday work practice in system design. Acta Univ Oul A 362. URL: http://herkules.oulu.fi/isbn9514259556/. Cited 2010/10/31. Käypä hoito (2006) Aivoinfarkti. Käypä hoito -suositus. URL: http://www.kaypahoito.fi/ web/kh/suositukset/naytaartikkeli/tunnus/hoi50051#s7. Cited 2010/10/31. Kim DK, Yoo SK & Kim SH (2005a) Instant wireless transmission of radiological images using a personal digital assistant phone for emergency teleconsultation. J Telemed Telecare 11(Suppl 2): S58–61. Kim YJ, Han JK, Kim SH, Jeong JY, An SK, Han CJ, Son KR, Lee KH, Lee JM & Choi BI (2005b) Small-bowel obstruction in a phantom model of ex vivo porcine intestine: comparison of PACS stack and tile modes for CT interpretation. Radiology 236(3): 867–871. Kinsey TV, Horton MC & Lewis TE (2000) Interfacing the PACS and the HIS: results of a 5-year implementation. Radiographics 20(3): 883–891. Kiuru A, Akisada M, Okabe T, Olsson S & Saranummi N (1991a) PACS in Japan and the Nordic Countries. First Japan-Nordic PACS symposium. Comput Methods Programs Biomed 36(2–3): 65–70. Kiuru A, Svedstrom E & Kuuluvainen I (1991b) Dynamic imaging of pulmonary ventilation. Description of a novel digital fluoroscopic system. Acta Radiol 32(2): 114–119. Kiuru A (1993) Kuvan ja kuvan oheistiedon standardointi. Suomen lääkärilehti 48(9): 827–831. Kiuru MJ, Paakkala TA, Kallio TT, Aalto J & Rajamaki M (2002) Effect of teleradiology on the diagnosis, treatment and prognosis of patients in a primary care centre. J Telemed Telecare 8(1): 25–31. Kiviaho K, Winblad I & Reponen J (2004) Terveydenhuollon toimintaprosesseja ja asiointia tukevat atk-sovellukset Suomessa – kartoitus ja käyttöanalyysi. Osaavien keskusten verkoston julkaisuja 8/2004. Koff D, Bak P, Brownrigg P, Hosseinzadeh D, Khademi A, Kiss A, Lepanto L, Michalak T, Shulman H & Volkening A (2009) Pan-Canadian evaluation of irreversible compression ratios ("lossy" compression) for development of national guidelines. J Digit Imaging 22(6): 569–578. Kondoh H & Okuda Y (2010) Monitor characteristics of iPhone/iPodtouch as a PACS viewer. Int J CARS 5(Suppl 1): S176–S177. Kormano M, Kallio T & Heinila J (1994) Teleradiology in Turku University Hospital-diagnostic evaluation and clinical applications. Comput Methods Programs Biomed 43(1–2): 125–128. Korsoff L, Kallio T, Kormano M & Heinila J (1995) Experiences with a teleradiology system in pulmonary diseases. Acta Radiol 36(1): 37–40. 109 Kotter E & Langer M (1998) Integrating HIS-RIS-PACS: the Freiburg experience. Eur Radiol 8(9): 1707–1718. Kotter E, Roesner A, Torsten Winterer J, Ghanem N, Einert A, Jaeger D, Uhrmeister P & Langer M (2003) Evaluation of Lossy data compression of chest X-rays: a receiver operating characteristic study. Invest Radiol 38(5): 243–249. Kovalerchuk B, Ruiz J, Vityaev E & Fisher S (1998) Prototype Internet consultation system for radiologists. J Digit Imaging 11(3 Suppl 1): 22–26. Kreutzer J, Akutsu H, Fahlbusch R, Buchfelder M & Nimsky C (2008) Teleradiology in neurosurgery: experience in 1024 cases. J Telemed Telecare 14(2): 67–70. Kroeker KI, Diamond R, Jennett PA & Johnston RV (2000) Video-capture teleradiology for the after-hours reading of computerized tomography scans. J Telemed Telecare 6(4): 229–232. Krupinski E, Nypaver M, Poropatich R, Ellis D, Safwat R & Sapci H (2002) Telemedicine/telehealth: an international perspective. Clinical applications in telemedicine/telehealth. Telemed J E Health 8(1): 13–34. Kumar S (2008) Introduction to Teleradiology. In Kumar S & Krupinski EA (eds) Teleradiology. Berlin Heidelberg, Spinger: 1–9. Kyriacou EC, Pattichis CS & Pattichis MS (2009) An overview of recent health care support systems for eEmergency and mHealth applications. Conf Proc IEEE Eng Med Biol Soc 2006: 1246–1249. L 159/2007 (2007) Laki sosiaali- ja terveydenhuollon asiakastietojen sähköisestä käsittelystä URL: http://www.finlex.fi/fi/laki/ajantasa/2007/20070159. Cited 2010/10/31. L 61/2007 (2007) Laki sähköisestä lääkemääräyksestä URL: http://www.finlex.fi/fi/laki/ ajantasa/2007/20070061. Cited 2010/10/31. L 629/2010 (2010) Laki terveydenhuollon laitteista ja tarvikkeista URL: http://www.finlex.fi/fi/laki/ajantasa/2010/20100629. Cited 2010/10/31. Läkemedelsverket (2009) The Medical Products Agency’s Working Group on Medical Information Systems. Project summary. Proposal for guidelines regarding classification of software based information systems used in health care. URL: http://www.lakemedelsverket.se/upload/foretag/medicinteknik/en/MedicalInformation-Systems-Report_2009-06-18.pdf. Cited 2010/10/31. Lamminen H (1999) Mobile satellite systems. J Telemed Telecare 5(2): 71–83. Langan JK & Wagner HN,Jr (1969) A display and storage system for radionuclide images. Radiology 93(5): 1197–1198. Lear JL, Manco-Johnson M, Feyerabend A, Anderson G & Robinson D (1989) Ultra-highspeed teleradiology with ISDN technology. Radiology 171(3): 862–863. Lee SK, Peng CH, Wen CH, Huang SK & Jiang WZ (1999) Consulting with radiologists outside the hospital by using Java. Radiographics 19(4): 1069–1075. Lee T, Latham J, Kerr RS, Kaar G, Teddy PJ, Dobson D & Anslow P (1990) Effect of a new computed tomographic image transfer system on management of referrals to a regional neurosurgical service. Lancet 336(8707): 101–103. 110 Lewis R, Jonathan H. Sunshine J & Bhargavan M (2009) Radiology Practices’ Use of External Off-Hours Teleradiology Services in 2007 and Changes Since 2003. AJR Am J Roentgenol. 193(5):1333–9. Ligier Y, Ratib O, Girard C, Rubin P & Reimer M (1993) A metropolitan Area Network for Teleradiology and Remote Expert Consultation Based on ISDN. Proceedings of CAR´92, Computer Assisted Radiology. Berlin Heidelberg, Springer: 112–117. Liukkonen E (2010) Radiologisten kuvien katselussa käytettävien näyttöjen laatu. Näyttöjen laitekanta, suorituskyky ja laadunvalvonta sekä kuvankatseluolosuhteet radiologisissa yksiköissä ja terveyskeskuksissa. Acta Univ Oul D 1058. URL: http://herkules.oulu.fi/isbn9789514262180/isbn9789514262180.pdf. Cited 2010/10/31. Lo SC & Huang HK (1986) Compression of radiological images with 512, 1,024, and 2,048 matrices. Radiology 161(2): 519–525. Logeswaran R (2008) Compression of Medical Images for Teleradiology. In Kumar S & Krupinski EA (eds) Teleradiology. Berlin-Heidelberg, Springer: 21–31. Loose R, Braunschweig R, Kotter E, Mildenberger P, Simmler R & Wucherer M (2009) Compression of digital images in radiology - results of a consensus conference. Rofo 181(1): 32–37. Lufkin RB, Wong WS, Winter J & Callisen HH (1983) Low-cost digital teleradiology systems. AJR Am J Roentgenol 140(2): 377–381. Lundberg N, Wintell M & Lindskold L (2010) The future progress of teleradiology -an empirical study in Sweden. Eur J Radiol 73(1): 10–19. MacMahon H, Doi K, Sanada S, Montner SM, Giger ML, Metz CE, Nakamori N, Yin FF, Xu XW & Yonekawa H (1991) Data compression: effect on diagnostic accuracy in digital chest radiography. Radiology 178(1): 175–179. Markivee CR & Chenoweth JL (1990) Teleradiology image transmission system: diagnostic accuracy at three matrix sizes for various types of images. J Digit Imaging 3(3): 170–173. Mathie AG & Strickland NH (1997) Interpretation of CT scans with PACS image display in stack mode. Radiology 203(1): 207–209. Microsoft Corporation (2009) Connected Health Framework Architecture and Design Blueprint. A Stable Foundation for Agile Health and Social Care. Part 1 – Introduction and Overview. URL: https://www.microsoft.com/industry/healthcare/ technology/HealthFramework.mspx. Cited 2010/10/31. Ministerial declaration 22/05/2003 (2003) Ministerial Conference Declaration: Ministers of EU Member States, Acceding and Associated countries, as well as EFTA countries on the development of national and regional eHealth implementation plans as an integral part of eEurope 2005 Action Plan. Adopted 22/05/2003. URL: http://ec.europa.eu/ information_society/eeurope/ehealth/conference/2003/doc/min_dec_22_may_03.pdf. Cited 2010/10/31. Ministerial declaration 15/03/2010 (2010) Ministerial Conference Declaration: European Co-operation on eHealth, Adopted 15/03/2010. URL: http://www.epractice.eu/files/ Final%20Conference%20Declaration%20-%20European%20Co-operation%20on%20 eHealth.pdf. Cited 2010/10/31. 111 Mun S, Elsayed A, Tohme W & Chris Wu Y (1995) Teleradiology/Telepathology requirements and implementation. J Med Syst 19(2): 153–164. Mun SK, Levine B, Cleary K & Dai H (1998) Deployable teleradiology and telemedicine for the US military. Comput Methods Programs Biomed 57(1–2): 21–27. Murphy RL,Jr & Bird KT (1974) Telediagnosis: a new community health resource. Observations on the feasibility of telediagnosis based on 1000 patient transactions. Am J Public Health 64(2): 113–119. Murphy RL, Barber D, Broadhurst A & Bird KT (1970) Microwave transmission of chest roentgenograms. Am Rev Respir Dis 102(5): 771–777. National Electrical Manufacturers Association (NEMA) (2010) DICOM - Strategic document, version 10.4. October 25, 2010. URL: http://medical.nema.org/dicom/geninfo/Strategy.pdf. Cited 2010/10/31. National Electrical Manufacturers Association (NEMA) (2009) Digital Imaging and Communications in Medicine (DICOM). Part 1: Introduction and Overview. 2009(05). URL: medical.nema.org/dicom/2004/04_01PU.PDF. Cited 2010/10/31. Niinimäki J, Holopainen A, Kerttula J & Reponen J (2001) Security development of a pocket-sized teleradiology consultation system. Stud Health Technol Inform 84(Pt 2): 1266–1270. Nilsson U, Nyman U & Nilsson M (1986) Teletransmission of radiographic images. Preliminary report. Acta Radiol Diagn (Stockh) 27(3): 357–360. Olsson S (1991) Picture archiving and communication systems and related developments in Sweden. J Digit Imaging 4(3): 131–136. Oosterwijk H (2004) Introduction to IHE. Int Congr Ser 1268: 92–95. O'Sullivan DC, Averch TD, Cadeddu JA, Moore RG, Beser N, Breitenbach C, Khazan R & Kavoussi LR (1997) Teleradiology in urology: comparison of digital image quality with original radiographic films to detect urinary calculi. J Urol 158(6): 2216–2220. Paakkala T, Aalto J, Kahara V & Seppanen S (1991) Diagnostic performance of a teleradiology system in primary health care. Comput Methods Programs Biomed 36(2–3): 157–160. Pagani L, Jyrkinen L, Niinimäki J, Reponen J, Karttunen A, Ilkko E & Jartti P (2003) A portable diagnostic workstation based on a Webpad: Implementation and evaluation. J Telemed Telecare 9(4): 225–229. Page G, Gregoire A, Galand C, Sylvestre J, Chahlaoui J, Fauteux P, Dussault R, Seguin R & Roberge FA (1981) Teleradiology in northern Quebec. Radiology 140(2): 361–366. Pattynama PM (2010) Legal aspects of cross-border teleradiology. Eur J Radiol 73(1): 26– 30. PDAdb.net (2005) Detailed Technical Specifications of O2 XDA (HTC Wallaby) URL: http://pdadb.net/index.php?m=specs&id=117&view=1&c=o2_xda_htc_wallaby. Cited 2010/10/31. PDAdb.net (2007a) Detailed Technical Specifications of Nokia 9000 Communicator. URL: http://pdadb.net/index.php?m=specs&id=879&view=1&c=nokia_9000_communicator. Cited 2010/10/31. 112 PDAdb.net (2007b) Detailed Technical Specifications of Nokia 9110 Communicator. URL: http://pdadb.net/index.php?m=specs&id=881&view=1&c=nokia_9110_communicator. Cited 2010/10/31. Plathow C, Walz M, Essig M, Engelmann U, Schulz-Ertner D, Delorme S & Kauczor HU (2005) Teleradiology: economic research analysis of CT investigations in a small hospital. Rofo 177(7): 1016–1026. Plowman L, Rogers Y & and Ramage M (1995) What are workplace studies for? In Marmolin H, Sundblad Y & Schmidt K (eds) Proceedings of the Fourth European Conference on Computer-Supported Cooperative Work. Dordrecht, Kluwer: 309–324. Poca MA, Sahuquillo J, Domenech P, Pedraza S, Maideu J, Vila X, Arikan F, Sanchez E & Garnacho A (2004) Use of teleradiology in the evaluation and management of headinjured patients. Results of a pilot study of a link between a district general hospital and a neurosurgical referral center. Neurocirugia (Astur) 15(1): 17–35. Pohjonen H, Kauppinen T & Ahovuo J (2004) ASP archiving solution of regional HUSpacs. Eur Radiol 14(9): 1702–1706. Protopapas Z, Siegel EL, Reiner BI, Pomerantz SM, Pickar ER, Wilson M & Hooper FJ (1996) Picture archiving and communication system training for physicians: lessons learned at the Baltimore VA Medical Center. J Digit Imaging 9(3): 131–136. Pyrros AT & Yaghmai Y (2008) Wireless Teleradiology and Security. In Kumar S & Krupinski E,A. (eds) Teleradiology. Springer Berlin Heidelberg, Springer: 227–241. Radiation and Nuclear Safety Authority (STUK) (2008) (2008) Terveydenhuollon röntgenlaitteiden laadunvalvontaopas. STUK tiedottaa 2/2008. URL: http://www.stuk.fi/julkaisut_maaraykset/fi_FI/stuk_tiedottaa/_files/122226325100222 73/default/STUK-tiedottaa-2-2008.pdf. Cited 2010/10/31. Raman B, Raman R, Raman L & Beaulieu CF (2004) Radiology on handheld devices: image display, manipulation, and PACS integration issues. Radiographics 24(1): 299– 310. Ratib O (2009) Imaging informatics: from image management to image navigation. Yearb Med Inform: 167–172. Reiner BI, Siegel EL, Hooper F, Pomerantz SM, Protopapas Z, Pickar E & Killewich L (1996) Picture archiving and communication systems and vascular surgery: clinical impressions and suggestions for improvement. J Digit Imaging 9(4): 167–171. Reponen J (1996) Teleradiologia Suomessa. Suom Lääkäril 51(17): 1875–1878. Reponen J (2004) Radiology as a part of a comprehensive telemedicine and eHealth network in Northern Finland. Int J Circumpolar Health 63(4): 429–435. Reponen J (2008) Teleradiology: A Northern Finland Perspective. In Kumar J & Krupinski EA (eds) Teleradiology. Berlin Heidelberg, Springer: 217–225. Reponen J, Marttila E, Paajanen H & Turula A (2004) Extending a multimedia medical record to a regional service with electronic referral and discharge letters. J Telemed Telecare 10(SUPPL. 1): 81–83. Reponen J & Niinimäki J (2006) Digitaalinen radiologia - modaliteeteista tietoverkkoihin. In Korhola O, Kivisaari L, Laasonen E, Laasonen L, Paakkala T & Tervonen O (eds) Radiologia Suomessa - historiikki vuoteen 2005. Helsinki, WSOY: 76–82. 113 Richter HP & Braun V (1993) Computerized tomography and craniocerebral trauma. Unfallchirurg 96(11): 576–581. Ricke J, Maass P, Lopez Hanninen E, Liebig T, Amthauer H, Stroszczynski C, Schauer W, Boskamp T & Wolf M (1998) Wavelet versus JPEG (Joint Photographic Expert Group) and fractal compression. Impact on the detection of low-contrast details in computed radiographs. Invest Radiol 33(8): 456–463. Roberson GH & Shieh YY (1998) Radiology information systems, picture archiving and communication systems, teleradiology--overview and design criteria. J Digit Imaging 11(4 Suppl 2): 2–7. Roine R, Ohinmaa A & Hailey D (2001) Assessing telemedicine: a systematic review of the literature. CMAJ 165(6): 765–771. Röntgen WC (1895) Ueber eine neue Art von Strahlen (vorläufige Mittheilung). Ber Phys Med Ges Wurzbg: 132–141. Ross P, Sepper R & Pohjonen H (2010) Cross-border teleradiology-experience from two international teleradiology projects. Eur J Radiol 73(1): 20–25. Rosset C, Rosset A & Ratib O (2005) General consumer communication tools for improved image management and communication in medicine. J Digit Imaging 18(4): 270–279. Ruotsalainen P (2010) Privacy and security in teleradiology. Eur J Radiol 73(1): 31–35. Salo S, Salo H, Liisanantti A & Reponen J (2007) Data transmission in dental identification of mass disaster victims. J Forensic Odontostomatol 25(1): 17–22. Samei E, Badano A, Chakraborty D, Compton K, Cornelius C, Corrigan K, Flynn MJ, Hemminger B, Hangiandreou N, Johnson J, Moxley M, Pavlicek W, Roehrig H, Rutz L, Shepard J, Uzenoff R, Wang J & Willis C (2005a) Assessment of Display Performance for Medical Imaging Systems, Report of the American Association of Physicists in Medicine (AAPM) Task Group 18. Madison, WI, AAPM On-Line Report No. 03, April 2005, Medical Physics Publishing. Samei E, Badano A, Chakraborty D, Compton K, Cornelius C, Corrigan K, Flynn MJ, Hemminger B, Hangiandreou N, Johnson J, Moxley-Stevens DM, Pavlicek W, Roehrig H, Rutz L, Shepard J, Uzenoff RA, Wang J & Willis CE (2005b) Assessment of display performance for medical imaging systems: executive summary of AAPM TG18 report. Med Phys 32(4): 1205–1225. Savcenko V, Erickson BJ, Palisson PM, Persons KR, Manduca A, Hartman TE, Harms GF & Brown LR (1998) Detection of subtle abnormalities on chest radiographs after irreversible compression. Radiology 206(3): 609–616. Savcenko V, Erickson BJ, Persons KR, Campeau NG, Huston J,3rd, Wood CP & Schreiner SA (2000) An evaluation of JPEG and JPEG 2000 irreversible compression algorithms applied to neurologic computed tomography and magnetic resonance images. Joint Photographic Experts Group. J Digit Imaging 13(2 Suppl 1): 183–185. Scarsbrook AF, Graham RN & Perriss RW (2005) The scope of educational resources for radiologists on the internet. Clin Radiol 60(5): 524–530. 114 Schulze RK, Richter A & d'Hoedt B (2008) The effect of wavelet and discrete cosine transform compression of digital radiographs on the detection of subtle proximal caries. ROC analysis. Caries Res 42(5): 334–339. Seibel RM, Kocher K & Landsberg P (2000) Security aspects on the Internet. Radiologe 40(4): 394–399. Shieh YY & Roberson GH (1999) Integrated radiology information system, picture archiving and communications system, and teleradiology--workflow-driven and future-proof. J Digit Imaging 12(2 Suppl 1): 199–200. Shieh YY, Tsai FY & Shieh M (2008) The impact of globalisation on teleradiology practice. Int J Electron Healthc 4(3–4): 290–298. Siegel E & Reiner B (2002) Work flow redesign: the key to success when using PACS. AJR Am J Roentgenol 178(3): 563–566. Siegel EL, Protopapas Z, Reiner BI & Pomerantz SM (1997) Patterns of utilization of computer workstations in a filmless environment and implications for current and future picture archiving and communication systems. J Digit Imaging 10(3 Suppl 1): 41–43. Slovis TL & Guzzardo-Dobson PR (1991) The clinical usefulness of teleradiology of neonates: expanded services without expanded staff. Pediatr Radiol 21(5): 333–335. Soegner P, Rettenbacher T, Smekal A, Buttinger K, Oefner B & zur Nedden D (2003a) Benefit for the patient of a teleradiology process certified to meet an international standard. J Telemed Telecare 9 Suppl 2: S61–2. Soegner P, Rettenbacher T, Smekal A & Zur Nedden D (2003b) Guidelines for teleradiology practice: results of the Tyrolean teleradiology pilot project. J Telemed Telecare 9(Suppl 1): S48–50. Soila P (1970) Luuranko kolisee kuvakaapissa. Tekniikan Maailma (14):82–85. Spencer JA, Dobson D, Hoare M, Molyneux AJ & Anslow PL (1991) The use of a computerized image transfer system linking a regional neuroradiology centre to its district hospitals. Clin Radiol 44(5): 342–344. Stanberry B (2006) Legal and ethical aspects of telemedicine. J Telemed Telecare 12(4): 166–75. Stewart BK, Dwyer SJ,3rd & Kangarloo H (1992) Design of a high-speed, high-resolution teleradiology network. J Digit Imaging 5(3): 144–155. STM (2009) Sosiaali- ja terveysministeriö (STM) Tiedote 385/2009 Ministeri Risikko: Terveydenhuollon sähköistä tiedonhallintaa vaiheistettava. URI: http://www.stm.fi/tiedotteet/tiedote/view/1438004#fi. Cited 2010/10/02. STM (2010a) Sosiaali- ja terveysministeriö (STM) Sosiaali- ja terveydenhuollon asiakastietojen sähköisestä käsittelystä annetun lain muuttaminen. URI: http://www.stm.fi/vireilla/lainsaadantohankkeet/sosiaali_ja_terveydenhuolto/sahkoiset. Cited 2010/10/02. STM (2010b) Sosiaali- terveysministeriö (STM) Tiedote 171/2010 Ensimmäiset sähköiset reseptit on kirjoitettu. URL: http://www.stm.fi/tiedotteet/tiedote/view/1508762#fi. Cited 2010/10/31. 115 Stoeger A, Strohmayr W, Giacomuzzi SM, Dessl A, Buchberger W & Jaschke W (1997) A cost analysis of an emergency computerized tomography teleradiology system. J Telemed Telecare 3(1): 35–39. Stoerkson SA & Aslaksen A (2010) Organisational, technical and professional reflections on radiological cross enterprise document sharing in the Western Norway Health care Region Perspectives of the innovation from an administrative and a radiological angle. Int J CARS 5(Suppl 1): S182. Stranzinger E, Treumann TC, Dreier D & Allgayer B (2003) Teleradiology for emergency cranial computed tomography. Rofo 175(5): 646–653. Strauss A, Fagerhaugh S, Suczek B & Wiener C (1985) Social organization of medical work. Chicago, University of Chicago. Struber J & Tichon J (2003) Minimum technical requirements and evaluating effectiveness of teleradiology. Technol Health Care 11(6): 391–398. Sung MM, Kim HJ, Yoo SK, Choi BW, Nam JE, Kim HS, Lee JH & Yoo HS (2002) Clinical evaluation of compression ratios using JPEG2000 on computed radiography chest images. J Digit Imaging 15(2): 78–83. Taira R, Mankovich N, Boechat M, Kangarloo H & Huang H (1988) Design and implementation of a picture archiving and communication system for pediatric radiology. Am J Roentgenol 150(5): 1117–1121. Tachakra S, Wang XH, Istepanian RH, Song YH (2003) Mobile e-Health: The Unwired Evolution of Telemedicine. Telemed J E Health 9(3): 247–257. Templeton AW, Dwyer SJ,3rd, Rosenthal SJ, Eckard DA, Harrison LA & Cook LT (1991) A dial-up digital teleradiology system: technical considerations and clinical experience. AJR Am J Roentgenol 157(6): 1331–1336. The Royal College of radiologists (RCR) (2008) The adoption of lossy data compression for the purpose of clinical interpretation. The Royal College of radiologists report, April 2008. URL: http://www.rcr.ac.uk/docs/radiology/pdf/ IT_guidance_LossyApr08.pdf. Cited 2010/10/31. Thrall JH (2005) Reinventing radiology in the digital age: part I. The all-digital department. Radiology 236(2): 382–385. Thrall JH (2007a) Teleradiology. Part I. History and clinical applications. Radiology 243(3): 613–617. Thrall JH (2007b) Teleradiology. Part II. Limitations, risks, and opportunities. Radiology 244(2): 325–328. Uldal SB, Sund T & Stormer J (1997) Four years with teleradiology: a technical description. Telemed J 3(3): 235–241. Vassallo DJ, Hoque F, Roberts MF, Patterson V, Swinfen P & Swinfen R (2001) An evaluation of the first year's experience with a low-cost telemedicine link in Bangladesh. J Telemed Telecare 7(3): 125–138. Viitanen J, Kiuru A, Heiska K, Kärkkäinen A & Kormano M (1991) Finnish PACS project. Comput Methods Programs Biomed 36(2–3): 99–102. 116 Viitanen J, Sund T, Rinde E, Stoermer J, Kormano M, Heinila J, Yliaho J & Ahonen J (1992) Nordic teleradiology development. Comput Methods Programs Biomed 37(4): 273–277. Vogl R (2005) Teleradiology and PACS--strategy of the Innsbruck University Hospital. Radiologe 45(8): 704–711. Voskarides S, Pattichis CS, Habib Istepanian RS, Kyriacou E, Pattichis MS, Schizas CN (2002) In: Rao RM, Dianat SA & Zoltowski (eds) Digital Wireless Communications IV. Proc SPIE 4740: 124–131. Vuorela A & Puranen M (1994) Teleradiologinen kuvansiirto nopeuttaa hoitopäätöksiä. Suom Lääkäril 49(11): 1181–1183. Waegemann CP (2010) mHealth: the next generation of telemedicine? Telemed J E Health 16(1): 23–25. Wagner HN,Jr (2002) The Internet: the road to more effective PET. Br J Radiol 75 Spec No: S20–3. Watanabe M, Jennett P & Watson M (1999) The effect of information technology on the physician workforce and health care in isolated communities: the Canadian picture. J Telemed Telecare 5 Suppl 2: S11–9. Wein B,B., Eichelberg M, Ihls A & Poiseau E (2005) IHE "Integrating the Healthcare Enterprise" -an update for Information Technology Infrastructure for 2005. Int Congr Ser 1281: 169–174. Wikipedia (2010) 3GPP Long Term Evolution. URL: http://en.wikipedia.org/wiki/ 3GPP_Long_Term_Evolution. Cited 2010/10/31. Wilson AJ (1996) Is teleradiology the solution to after-hours emergency radiology coverage? Radiographics 16(4): 939–942. Winblad I, Hämäläinen P & Reponen J (2010) Kansallinen terveysarkisto - vajaa vuosi aikaa liittymiseen. Osa I : mistä uudessa tietojärjestelmässä on kyse? Suom Lääkäril 65,(13–14): 1218–1222. Winblad I, Reponen J, Hämäläinen P & Kangas M (2006) Informaatio- ja kommunikaatioteknologian käyttö Suomen terveydenhuollossa vuonna 2005. Tilanne ja kehityksen suunta. Raportteja 7/2006, STAKES, Helsinki. Wong WS, Roubal I, Jackson DB, Paik WN & Wong VK (2005) Outsourced teleradiology imaging services: an analysis of discordant interpretation in 124,870 cases. J Am Coll Radiol 2(6): 478–484. Wootton R (1996) Telemedicine: a cautious welcome. BMJ 313(7069): 1375–1377. Wootton R (2001) Recent advances: Telemedicine. BMJ 323(7312): 557–560. Wootton R, Youngberry K, Swinfen R & Swinfen P (2005) Referral patterns in a global store-and-forward telemedicine system. J Telemed Telecare 11 Suppl 2: S100–3. World Health Organization (2010) http://www.who.int/topics/ehealth/en/. 2010(05/24). Wunderbaldinger P, Schima W, Turetschek K, Helbich TH, Bankier AA & Herold CJ (1999) World Wide Web and Internet: applications for radiologists. Eur Radiol 9(6): 1170–1182. 117 Yaghmai V, Salehi SA, Kuppuswami S & Berlin JW (2004) Rapid wireless transmission of head CT images to a personal digital assistant for remote consultation. Acad Radiol 11(11): 1291–1293. Yamamoto LG (1995) Wireless teleradiology and fax using cellular phones and notebook PCs for instant access to consultants. Am J Emerg Med 13(2): 184–187. Yamamoto LG, Ash KM, Boychuk RB, Balaraman V, Iwaishi LK, Sommer-Candelario SA, Amell CM & Bennett MI (1996) Personal computer teleradiology interhospital image transmission of neonatal radiographs to facilitate tertiary neonatology telephone consultation and patient transfer. J Perinatol 16(4): 292–298. Yamamoto LG, DiMauro R & Long DC (1993) Personal computer teleradiology: comparing image quality of lateral cervical spine radiographs with conventional teleradiology. Am J Emerg Med 11(4): 384–389. Yamamoto LG, Wiebe RA, Long DC & Dimauro R (1992) Scanned computed tomography image transmission by modem: feasibility for emergency department transfers. Am J Emerg Med 10(3): 226–229. Yamamoto LG & Williams DR (2000) A demonstration of instant pocket wireless CT teleradiology to facilitate stat neurosurgical consultation and future telemedicine implications. Am J Emerg Med 18(4): 423–426. Yoho DR,Jr (1994) Wireless communication technology applied to disaster response. Aviat Space Environ Med 65(9): 839–845. Zhang J, Yu F, Sun J, Yang Y, Liang C (2007) DICOM image secure communications with Internet protocols IPv6 and IPv4. IEEE Trans Inf Technol Biomed 11(1): 70–80. 118 Original publications This thesis is based on the following original articles, which are referred to in the text by their Roman numerals. I II III IV V VI VII Reponen J, Lähde S, Tervonen O, Ilkko E, Rissanen T & Suramo I (1995) Low-cost Digital Teleradiology. Eur J Radiol 19: 226–231. Karasti H, Reponen J, Tervonen O & Kuutti K (1998) The teleradiology system and changes in work practices. Comput Methods Programs Biomed 57(1–2): 69–78. Reponen J, Palsson T, Kjartansson O, Ilkko E, Störmer J, Päivänsalo M, Pyhtinen J, Laitinen J, Eiriksson A & Brekkan A (1995) Nordic Telemedicine Consultation Network Using Internet. Proceedings of CAR ´95 Computer Assisted Radiology and Surgery. Berlin, Springer Verlag: 723–728. Reponen J, Ilkko E, Jyrkinen L, Karhula V, Tervonen O, Laitinen J, Leisti E-L, Koivula A & Suramo I (1998) Wireless radiological consultation with portable computers. J Telemed Telecare 4: 201–205. Reponen J, Ilkko E, Jyrkinen L, Tervonen O, Niinimaki J, Karhula V & Koivula A (2000) Initial experience with a wireless personal digital assistant as a teleradiology terminal for reporting emergency computerized tomography scans. J Telemed Telecare 6(1): 45–9. Reponen J, Niinimäki J, Kumpulainen T, Ilkko E, Karttunen A & Jartti P (2005) Mobile teleradiology with smartphone terminals as a part of a multimedia electronic patient record. Int Congr Ser 1281: 916–921. Reponen J, Winblad I & Hämäläinen P (2008) Current status of national eHealth and telemedicine development in Finland. Stud Health Technol Inform 134: 199–208. Reprinted with permission from Elsevier (I, II, VI), Springer (III), Royal Society of Medicine (RSM) press (IV, V) and IOS press (VII). Original publications are not included in the electronic version of the thesis. 119 120 ACTA UNIVERSITATIS OULUENSIS SERIES D MEDICA 1063. Gäddnäs, Fiia (2010) Insights into healing response in severe sepsis from a connective tissue perspective : a longitudinal case-control study on wound healing, collagen synthesis and degradation, and matrix metalloproteinases in patients with severe sepsis 1064. Alaräisänen, Antti (2010) Risk factors and pathways leading to suicide with special focus in schizophrenia : the Northern Finland 1966 Birth Cohort Study 1065. Oja, Paula (2010) Significance of customer feedback : an analysis of customer feedback data in a university hospital laboratory 1066. Mustonen, Erja (2010) The expression of novel, load-induced extracellular matrix modulating factors in cardiac remodeling 1067. Mäkinen, Johanna (2010) Systematic search and evaluation of published scientific research : implications for schizophrenia research 1068. Jartti, Pekka (2010) Computed tomography in subarachnoid haemorrhage : studies on aneurysm localization, hydrocephalus and early rebleeding 1069. Rantala, Aino (2010) Susceptibility to respiratory tract infections in young men: the role of inflammation, mannose-binding lectin, interleukin-6 and their genetic polymorphisms 1070. Tanskanen, Päivikki (2010) Brain MRI in subjects with schizophrenia and in adults born prematurely : the Northern Finland 1966 Birth Cohort Study 1071. Abass, Khaled M. (2010) Metabolism and interactions of pesticides in human and animal in vitro hepatic models 1072. Luukkonen, Anu-Helmi (2010) Bullying behaviour in relation to psychiatric disorders, suicidality and criminal offences : a study of under-age adolescent inpatients in Northern Finland 1073. Ahola, Riikka (2010) Measurement of bone exercise : osteogenic features of loading 1074. Krüger, Johanna (2010) Molecular genetics of early-onset Alzheimer's disease and frontotemporal lobar degeneration 1075. Kinnunen, Urpo (2010) Blood culture findings during neutropenia in adult patients with acute myeloid leukaemia : the influence of the phase of the disease, chemotherapy and the blood culture systems 1076. Saarela, Ville (2010) Stereometric parameters of the Heidelberg Retina Tomograph in the follow-up of glaucoma Book orders: Granum: Virtual book store http://granum.uta.fi/granum/ D 1077 OULU 2010 U N I V E R S I T Y O F O U L U P. O. B . 7 5 0 0 F I - 9 0 0 1 4 U N I V E R S I T Y O F O U L U F I N L A N D U N I V E R S I TAT I S S E R I E S SCIENTIAE RERUM NATURALIUM Professor Mikko Siponen HUMANIORA University Lecturer Elise Kärkkäinen TECHNICA Professor Hannu Heusala MEDICA Professor Olli Vuolteenaho SCIENTIAE RERUM SOCIALIUM ACTA TELERADIOLOGY— CHANGING RADIOLOGICAL SERVICE PROCESSES FROM LOCAL TO REGIONAL, INTERNATIONAL AND MOBILE ENVIRONMENT Senior Researcher Eila Estola SCRIPTA ACADEMICA Information officer Tiina Pistokoski OECONOMICA University Lecturer Seppo Eriksson EDITOR IN CHIEF Professor Olli Vuolteenaho PUBLICATIONS EDITOR Publications Editor Kirsti Nurkkala ISBN 978-951-42-6371-2 (Paperback) ISBN 978-951-42-6372-9 (PDF) ISSN 0355-3221 (Print) ISSN 1796-2234 (Online) U N I V E R S I T AT I S O U L U E N S I S Jarmo Reponen E D I T O R S Jarmo Reponen A B C D E F G O U L U E N S I S ACTA A C TA D 1077 UNIVERSITY OF OULU, FACULTY OF MEDICINE, INSTITUTE OF DIAGNOSTICS, DEPARTMENT OF DIAGNOSTIC RADIOLOGY; UNIVERSITY OF OULU, FACULTY OF MEDICINE, INSTITUTE OF BIOMEDICINE, DEPARTMENT OF MEDICAL TECHNOLOGY D MEDICA