Download 5_Characteristics of a Hybrid Image Archiving and Communication

349
Original
Characteristics of a Hybrid Image Archiving and
Communication System Serving a NeurologicalNeurosurgical Facility
Características de um Sistema de Comunicação e Arquivamento de Imagens Híbrido em um
Serviço de Neurologia-Neurocirurgia
Carolina Martins1
Thalita Santos2
Anna Craig-McQuaide3
Luiz Aguiar4
Marcelo Valença5
Valdenice Melo6
Gilberto Falbo7
Caio Leão8
ABSTRACT
The Clinical Images Database (BIC) is an archive of clinical images for medical care and research. Objective: To describe the
characteristics of and requirements for the creation of a BIC in a tertiary center in Northeastern Brazil. Methods: Technological
advances allow for long-term storage of clinical images and retrieval according to clinical needs. With the patient consent
and image anonymousness, the database can be used for research. It allows measurements/comparisons, reconstruction/
deconstruction and analysis of anatomical, clinical and pathological variables. The specific infrastructure and image capture
and storage requirements for a BIC used for clinical and research purposes are analyzed. Results: Image capture and storage:
Comprises centralization of X-ray, CT, MRI and angiography data through the network to a server. An anonymous version of
each set of studies must be available for use in research. External backup is needed. Visualization: Comprises the user interface
and includes stations for viewing and data processing. Maintenance: A technical team assists users and manages the system,
troubleshooting problems, registering users and performing regular audits. An external committee supervised the creation of
the BIC, and continues to review its progress. Conclusions: The present study could serve as a basis for the creation of similar
clinical databases for research use.
Key words: Health Information Systems; Medical imaging; PACS; Picture archiving and communication system
RESUMO
A Instituição na qual o estudo foi realizado é uma unidade terciária, especializada, única no SUS em perfil cardio-neuro. O Banco
de imagens clínicas (BIC) é um acervo de imagens para assistência e pesquisa. Objetivo: Descrever esferas da criação do BIC
em um Hospital do Nordeste do Brasil. Material e Métodos: Tecnologia da informação e softwares permitem armazenamento
de imagens clínicas garantindo longevidade aos dados de pacientes e acervo para estudos. Essas imagens podem oferecer
medidas/comparações, reconstrução/desconstrução, análise de aspectos anatômicos, clínicos, patológicos e antropológicos.
Necessidades específicas (infra-estrutura, captação, visualização, armazenamento de imagens) em um sistema que atenda
assistência e pesquisa são analisadas. Resultados: 1) Captação/armazenamento: Compreende centralização dos dados de CRraios-X, tomógrafo, ressonância e hemodinâmica, em rede, ao servidor. Backup simultâneo: fora das dependências da Instituição;
2) Visualização/tratamento: Compreende 40 estações incluindo visualização e tratamento de dados; 3) Manutenção/auditoria:
Uma equipe técnica acompanha o sistema, soluciona problemas, cadastra usuários e realiza auditoria periódica. A instalação
do BIC é acompanhada por Comitê de membros externos à instituição que acompanha instalação e funcionamento do acervo.
Conclusões: Esse acervo é de interesse a diversas áreas e instituições, podendo servir de base à criação de outras coleções de
interesse científico.
Palavras-chave: Imagens médicas; PACS; Sistema de arquivamento e distribuição de imagens; Sistemas de informação em saúde
Neurocirurgiã e Doutora em Anatomia e Cirurgia – FMRP-USP, Diretora de Ensino e Pesquisa do Hospital Pelópidas Silveira - HPS/IMIP/SES/SUS – Recife, PE, Brasil.
Enfermeira e Técnica em Pesquisa do Banco de Imagens Clínicas – HPS/IMIP/SES/SUS Recife, PE, Brasil.
3
Estagiária em Neurocirurgia, Imperial College London – Londres, Reino Unido.
4
Professor Titular do Programa de Pós-Graduação em Tecnologia em Saúde- Pontifícia Universidade Católica do Paraná – PUC-PR, Brasil
5
Professor Livre-Docente, Pós Graduação em Neurociências -UFPE –Recife, PE, Brasil
6
Enfermeira, Mestre em Neurociências - UFPE, Hospital Pelópidas Silveira-IMIP/SES/SUS–Recife, PE, Brasil
7
Diretor Administrativo, Hospital Pelópidas Silveira-IMIP/SES/SUS–Recife, PE, Brasil
8
Neurocirurgião e Superintendente, Hospital Pelópidas Silveira-IMIP/SES/SUS – Recife, PE, Brasil
1
2
Received May 29, 2015. Accepted Jan 04, 2016
Martins C, Santos T, Craig-McQuaide A, Aguiar L, Valença M, Melo V, Falbo G, Leão C. - Characteristics of a
Hybrid Image Archiving and Communication System Serving a Neurological-Neurosurgical Facility
J Bras Neurocirurg 25 (4): 349 - 354, 2014
350
Original
I ntroduction
The Bank of Clinical Images (BIC) is a collection of medical
images which can be used for clinical and research purposes.
Various institutions in Brazil use clinical image databases
in PACS (Picture Archive and Communication System)
format5,6,11,12. Hybrid image databases used for both clinical
care and medical research exist in Europe, the USA and
Japan8-10. The storage and use of data in hybrid image databases
must comply with medical ethics regulations.
The institution which hosts the BIC is a tertiary referral center
specialized in neurosurgery, neurology and cardiology. It is
a unique center to exist in the Brazilian public health system
(SUS – Sistema Único de Saúde). It has approximately a dozen
Digital Imaging and Communications in Medicine (DICOM)
imaging units generating a volume of 45,000 images per year.
This study aims to: a) describe the features of the BIC
hybrid system, including physical infrastructure, number and
characteristics of the imaging units, network features, and data
logging system; b) describe the features of the software used
to manage the image database, data visualization framework
for care and research, anonymousness method, user type; c)
compare the details of the BIC operating in the institution with
other existing models in the literature.
Methods
Compilation of the data used in the construction of the
hospital´s dual purpose image database system (BIC), which
are presented in descriptive and tabular forms. We performed
a systematic literature review and comparison of our data with
other systems which are described in the literature.
R esults
General Characteristics
The Bank of Clinical Images (BIC) is a medical image collection
obtained from the integration of the PACS with Radiology
Information Systems (RIS) and Hospital Information System
Martins C, Santos T, Craig-McQuaide A, Aguiar L, Valença M, Melo V, Falbo G, Leão C. - Characteristics of a Hybrid
Image Archiving and Communication System Serving a Neurological-Neurosurgical Facility
(HIS). It comprises the acquisition, processing and archiving
of digital images in DICOM standard, enabling the sharing of
information through local or remote access.
The Hospital that hosts the BIC is the only hospital in
the Brazilian public health system (SUS) specialized in
neurosurgery, neurology and cardiology. As such, it has the
potential to provide a basis for cutting-edge research in each
of these areas. Opened in 2011, the hospital has 200 beds, of
which 30 in intensive care, 60 neurosurgical beds, 30 beds
each for cardiology and neurology, in addition to those in the
emergency department and the post-operative recovery room.
The institution has a number of digital imaging tools in the
DICOM standard (Table 1).
The Hospital Pelópidas Silveira Image Database (BIC-HPS)
consists of a main server and an associated server to run the
system and 20 terabytes of storage. Seven DICOM imaging
modalities, including computed tomography (CT), ultrasound
(US), and angiography are stored on the system. BIC-HPS
performs on average 3,750 imaging studies per month,
amounting to 408,000 images per month, and generating a total
output of 304 gigabytes per month. The estimated total storage
capacity is predicted to meet hospital demand for three to five
years. Other imaging modalities will be added to the system in
the near future.
At present, BIC-HPS houses 6,809 clinical studies. Of these
there are 4.678 CT and computed tomography angiogragram
(CTA), 1,023 are ultrasound images including echocardiogram
and Doppler studies, 1,078 are angiograms and 30 are
fluoroscopic images or intraoperative imaging. Two hundred
and thirty five (235) active users access the acquired images
for clinical and research purposes via forty (40) workstations
in the hospital. Remote access can be provided for registered
researchers working in other locations.
BIC-HPS is maintained by information technology (IT) and
healthcare professionals, who perform system management
and troubleshooting, monitor the use of applications, provide
training and perform monthly audit. This imaging archive
is monitored by an administrative committee external to
the institution and complies with ethical regulations. The
distribution of the image types taken in the hospital can be
found in Table 2, and includes X-ray, computed tomography,
magnetic resonance imaging (MRI), ultrasound, and cardiac
and cerebral angiograms.
J Bras Neurocirurg 25 (4): 349 - 354, 2014
351
Original
The BIC configuration characteristics are shown in Tables 1 to
3, and Figure 1.
Figure 1. BIC structure.
Discussion
In the clinical context, an image database is important in order
to guarantee the long-term storage of radiological information
obtained for each patient, in addition to optimizing patient
management through the sharing of information via the local
area network (intranet). Another feature of the PACS system
relevant to its clinical use is the ability to manipulate and
visualize medical images without the use of conventional
films2,7,12. This reduces significantly the maintenance costs6.
In terms of its research applications, the images in the database
can be accessed by internal and external researchers, providing
tools for knowledge generation in all areas of healthcare. To
this end, the images are anonymous and stored systematically,
constituting a collection for audit and research, which is
available to researchers locally as well as for remote access.
The main purpose of the BIC is to be a tool for clinical care
and for research. Therefore its set up works as a hybrid dualpurpose image database, combining the advantages described
above.
Hybrid image databases are commonly used in countries
including France, the USA and Japan8-10, which are the
forefront of medical technology. In Brazil, although the storage
of medical images for research purposes is not universal, such
systems do exist, mainly in Southern and Southeastern regions,
including Curitiba, Porto Alegre, Ribeirão Preto and São
Paulo5,6,11,12. In some of these locations, PACS systems were
implemented thanks to the development of specific software
which met the needs of the database, thus reducing operating
costs1,3,4 (Table 4a and 4b).
Martins C, Santos T, Craig-McQuaide A, Aguiar L, Valença M, Melo V, Falbo G, Leão C. - Characteristics of a
Hybrid Image Archiving and Communication System Serving a Neurological-Neurosurgical Facility
J Bras Neurocirurg 25 (4): 349 - 354, 2014
352
Original
System Architecture
The architecture of a hybrid image database, which integrates
DICOM images obtained from various terminals and systems,
is shown in Figure 1.
The DICOM images are centralized via network, archived
on a server with a pre-determined capacity, according to the
expected volume of investigations. This server is coupled to
a hard disk backup storage. These images are linked to each
patient, via his/her hospital registration number, creating
unique files for each patient and preventing duplication.
Data recovery in a clinical PACS system requires management
software (Table 1) working with the same coding as the
hospital information system, ensuring that all images relating
to a patient can be retrieved at any time. In a dual purpose
image database such as BIC, all images must be anonymous
in order to be used for research. Anonymousness is achieved
without loss of internal linkage, ensuring that past and present
image studies for each patient are linked, and allowing the
user to consult an up-to-date image collection which does not
contain any patient identifiable information.
The architecture of a dual purpose image database system
involves the development of a security system necessary for
a hospital environment and supported by servers designed
for the estimated demand. The network used by the system,
characterized in Table 3, should have adequate capacity for the
distribution of stored images via teleradiology.
Table 4a: Comparison between requerements for the installation of PACS according
to the published literacture and basic requirements for BIC.
Accessing a PACS system, solely for clinical use or dualpurpose, requires infrastructure comprising terminals, or
workstations, which allow registered users to access images
using a username and password. Access terminals can be
designated for clinical care and/or research. For both, the
programs used must be certified by the main international
healthcare safety authorities, because the data will be used
clinically, either directly or indirectly. Access terminals
for clinical use, which are more numerous, are distributed
according to service need. These terminals can have a simpler
configuration and work with basic programs for displaying
images, with limited image interpretation resources, as their use
is chiefly clinical. The workstations used for research require
higher specifications, including basic and advanced programs
for image display, and also for calculations, reconstructions
and advanced interpretation.
For the clinical image database which is currently being
developed, workstations dedicated to research are gathered in
the hospital library. This area can accommodate up to twentyfive users at any time, and has twelve computer terminals. The
library has a collection of approximately six hundred titles in
the fields of cardiology and neurology, and online access to
national and international journals.
Martins C, Santos T, Craig-McQuaide A, Aguiar L, Valença M, Melo V, Falbo G, Leão C. - Characteristics of a Hybrid
Image Archiving and Communication System Serving a Neurological-Neurosurgical Facility
J Bras Neurocirurg 25 (4): 349 - 354, 2014
353
Original
Table 4b: Comparison between requerements for the installation of PACS according
to the published literacture and basic requirements for BIC.
R eferences
1.
Azevedo-Marques PM, Caritá EC, Benedicto AA, Sanches PR.
Integração RIS/PACS no Hospital das Clínicas de Ribeirão Preto:
uma solução baseada em “web”. Radiol Bras. 2005;38(1):37-43.
http://dx.doi.org/10.1590/S0100-39842005000100009
2. Azevedo-Marques PM, Trad CS, Elias Junior J, Santos AC.
Implantação de um mini-pacs (sistema de arquivamento e
distribuição de imagens) em hospital universitário. Radiol Bras.
2001;34(4):221-4
Conclusion
Defining the essential features for the creation of a dual purpose
image database with a budget commensurate with a Brazilian
public healthcare system (SUS) facility is a pioneering
initiative and of scientific interest.
The construction of a hybrid image database or the systematic
association of the search interface to an existing image bank,
opens unprecedented opportunities for research in various
areas, including medical and human sciences, computer
science and information transmission, clinical engineering, and
the creation of new healthcare information systems. Because it
ensures digital data storage, a hybrid clinical image database
allows information to be accessed locally or remotely, which
favors the use of the collection by external researchers.
Martins C, Santos T, Craig-McQuaide A, Aguiar L, Valença M, Melo V, Falbo G, Leão C. - Characteristics of a
Hybrid Image Archiving and Communication System Serving a Neurological-Neurosurgical Facility
3.
Caritá EC, Matos ALM, Azevedo Marques PM. Ferramentas
para visualização de imagens médicas em hospital universitário.
Radiol Bras. 2004;37(6):437-40. http://dx.doi.org/10.1590/
S0100-39842004000600010
4.
Caritá EC, Seraphim E, Honda MO, Azevedo Marques PM.
Implementação e avaliação de um sistema de gerenciamento
de imagens médicas com suporte à recuperação baseada
em conteúdo. Radiol Bras. 2008;41(5):331-6. http://dx.doi.
org/10.1590/S0100-39842008000500012.
5.
Cotrim DS, Marques da Silva AM, Bezerra EA. Infra-estrutura
de informática para sistemas de apoio ao diagnóstico aplicada
a servidores PACS; 2007, 7 a 9 de novembro; III Simpósio de
Instrumentação e Imagens Médicas – SIIM; Brasil, São Carlos;
2007
6.
Galina CP, Gambarato VTS. Descrição da implantação do PACS
(Picture Archiving and Communication System) em um hospital
escola para redução de custos operacionais. Tékhne & Logos.
2012;3:1-20
7.
Henrique Neto G, Oliveira W, Valeri FV. Armazenamento de
Imagens Médicas com InterBase. Infocomp Journal of Computer
Science. 2004;3(1):13-17
8.
Huang HK. Short history of PACS. Part I: USA. Eur J Radiol.
2011;78(2):163-76. doi: 10.1016/j.ejrad.2010.05.007.
9.
Inamura K, Kim JH. History of PACS in Asia. Eur J Radiol.
2011;78(2):184-9. doi: 10.1016/j.ejrad.2010.09.022.
10. Lemke HU. Short history of PACS (Part II: Europe). Eur J
Radiol. 2011;78(2):177-83. doi: 10.1016/j.ejrad.2010.05.031.
11. Oliveira AF, Lederman HM, Batista NA. O aprendizado
sobre a tecnologia no diagnóstico por imagem. Radiol Bras.
2014;47(1):18-22.
12. Salomão SC, Azevedo Marques PM. Integrando ferramentas
de auxílio ao diagnóstico no sistema de arquivamento e
comunicação de imagens. Radiol Bras. 2011;44(6):374-380.
http://dx.doi.org/10.1590/S0100-39842011000600009.
J Bras Neurocirurg 25 (4): 349 - 354, 2014
354
Original
corresPondinG Author
Dra. Carolina Martins
Diretora de Ensino e Pesquisa
Hospital Metropolitano Oeste Pelópidas
Silveira IMIP/SES/SUS
Rodovia BR 232, Km 06, Curado – Recife, PE
Fone: +55 (081) 31830568
E-mail: [email protected]
Research carried out at Hospital Metropolitano
Oeste Pelópidas Silveira – IMIP/ SES/ SUS.
Funded by: Fundação de Amparo à Ciência e
Tecnologia do Estado de Pernambuco FACEPE.
Martins C, Santos T, Craig-McQuaide A, Aguiar L, Valença M, Melo V, Falbo G, Leão C. - Characteristics of a Hybrid
Image Archiving and Communication System Serving a Neurological-Neurosurgical Facility
J Bras Neurocirurg 25 (4): 349 - 354, 2014