Download A Web Based Medical Image Data Processing and Management

A Web Based Medical Image Data Processing and Management System
Jinman Kim1, David D. Feng1,2, Tom Weidong Cai1,2
1.
Biomedical and Multimedia Information Technology (BMIT) Group,
Basser Department of Computer Science, The University of Sydney, Australia
2.
Center for Multimedia Signal Processing (CMSP)
Department of Electronic & Information Engineering, Hong Kong Polytechnic University
Abstract
Medical services in current-time rely heavily on digital
imaging technology due to image modalities utilized in
medical field such as the computer tomography and magnetic
resonance imaging. These techniques require imageprocessing tools and digital management has been the primary
reason for development of Picture Archiving and Communication Systems (PACS). The limitation to PACS is that the
system architecture limits itself to be only used as a local area
network, which at times causes inconvenience towards
physicians requiring the system in the outside environment.
With the rapid advances in the network speed, computer
processing capabilities, and the support for online software
development, the potential to bridge the gap between the
hospital and outside environment to share and work in
conjunction may be realized.
In this paper, we present an application of database and
functional imaging that runs through the Internet network
(WWW) and Internet browser that has similar ability to the
PACS yet having the advantages of being an online system.
Keywords: World Wide Web (WWW), PACS, Java Applet,
Medical Image, and Image Processing.
I. INTRODUCTION
Recent advances in imaging technology allow evaluation of
medical images in depth, which are now the requirement when
studying and diagnosing an image. This depth in medical
imaging is the image processing, techniques that manipulate
the image to transform, enhance, and restore information. It is
the technique of improving information for human
interpretation, and processing of scene data for autonomous
machine perception. Until recently the possibility of studying
and diagnosing of medical images beyond the dedicated
workstations within the working department has not been fully
exposed that restricted the usage and convenience.
Current network technology allows for image processing
and data retrieval efficiently, reliably, and economically.
Allowance of such ability to be accessible online with the
usage of standard Internet network and browser can be of a
great beneficial towards medical imaging.
Past researches in the field of online imaging has produced
many researches into this field. The ability to retrieve medical
data and basic image processing on the Internet has shown
great potential and functionality. For instance, Slomka, Elliot,
and Driedger[3] developed a system that is capable of online
database retrieval and image processing of nuclear medical
imaging using the Internet browser and the World Wide Web.
Medical imaging plays a significant role in the study and
understanding of patient’s diagnostic examinations. The
demand for online medical imaging systems that allow
visualization and processing has increased significantly. These
systems allow the tools for visualization and image processing
to be accessible outside the working environment of the
physicians and also accessible at all times and in any location.
The aim of this research was to explore the potential of the
Internet and the browser technology, and to maximize the use
of the browser to eliminate the requirement of the dedicated
workstation in the field of medical imaging while still
maintaining the image processing functions and imaging tools.
Incorporated with this aim for the research has been the goal
to ensure that the system being developed will actually be
usable and functional to the potential user audience of the
system.
This study attempts to address the above issues and provide
a solution, which is the construction of the application
‘Functional Imaging Web’ (FIWeb).
II. SYSTEM ARCHITECTURE AND FUNCTIONALITY
A. System Components
The system architecture shown in Figure 1 comprises of
five components that are namely; Active Database, Graphical
User Interface, Image processing, Image Analysis, and
Imaging tools. There are two advantages associated with this
architecture. First, it is flexible in the database structure where
the system can be used with any kind of database setup such as
the SQL system or the text-based system (the one used in the
FIWeb).
Second, the system is designed to be extensible so that new
component may be added to the system to enhance and
customize the system. In the following, each component
together with how they interact with each other to deliver the
functionality is briefly described.
Staff – having access to all the patient information, Patient –
ability to view their own records, and Guest – guest user to
evaluate the system. Depending on the access level, there are
restrictions on the usage of the functions to the database. For
example, the patient access can not create/delete any accounts
and also are restricted to allow searching of their own data set
whereas Staff access can create/delete and search all the
patient information. Others functions that are available is the
‘view log’ for administrator user that records all the access and
events that has occurred in the database. Also the ability to
change the password of the user is implemented for further
security.
Figure 1. System architecture displaying major the
component of the FIWeb.
The ‘graphical user interface’ (GUI) is the component that
allows the interaction of the users to the system. The GUI is
possible by the use of the browser and the Java Applet, a
software-developing tool that allows for on-line development.
‘Database’ is the database setup that is used by the system to
store and retrieve medical information. It allows for data
management that is beneficial in searching, retrieval and
modifying.
Component ‘Region of Interest’, ‘Image processing’,
‘Image analysis’ and ‘Imaging tools’ are used to manipulate
and extract information from images in study or diagnosis.
These components along with features of the database are
explained below.
B. System functionality
1.) Patient Information and Image Database
The database developed is web-based database using
JavaScript, Python script, and HTML that can be used on-line
with any Internet browser that supports scripting language. It
uses python (ver 6.1) script language on the server side to
execute all the user commands such as login service and data
search and data retrieval. The security of the users is assured
by using the 16-bit rotor encryption to store sensitive data
from tempering and illegal access. The database allows
searching by several key words such as the patient name,
study type, and study date. Database also allows for creation
of accounts of the users and the patient.
There are access levels between the users. Four levels exist
that are the Administrator – maintaining the database system,
Figure 2. Sample window of the result from the Database
query searching.
2.) Functional Imaging System
FIWeb applet applications include beneficial medical
image processing techniques and tools. These consist of the
following.
 Region of Interest (ROI)
ROI is to select a region of an image and study that
segment of the image further. The user can make
selection by drawing a circle, rectangle, or a shape on the
image. There is no loss of information on the process of
ROI and all of the image’s attributes are maintained.
Once the region is selected, it is cropped and scaled
(optional) before being displayed as a new image. The
cropping of the image is archived by the masking of the
pixels of the image, setting every pixel to be true or false
(true being the pixel in the selected region) before
removing the false pixels. (Fig.4)

Image analysis
Analyzing of an image is a requirement when studying to
extract details such as pixel count, pixel mean, standard
deviation, color bin selection, and histogram. These
statistical informations are available in the FIWeb.
Figure 3. Functional Imaging Web, showing the main User Interface and some of the features like the Cine player, Image
selection panel, Image transformation, and the ROI.

Imaging tools
Cine player (sequence of image displayed at variable
frame rate) can be used to view the image data set as a
movie. Comment tool allows for users to note down
study results that can be loaded and appended at another
time. Compare operator which is used to visually
compare two images by fusing two images using methods
such as ‘XOR’, ‘OR’, ‘AND’ and ‘SUBTRACT’.

Image processing
FIWeb include common medical imaging operation such
as ‘interactive color table manipulation’, ‘Blur/Sharp’,
‘Color filter’, ‘Upper/Lower bound’ and ‘Median (3x3,
5x5, and 7x7 kernel) filtering’. Transformation methods
are also present with the ability to Scale, Rotate and
Translate.
magic’ that are for image processing. Hence the design of the
FIWeb has been the product of extracting interface design
from such applications and applying on to the medical
imaging. FIWeb conforms to these standard specifications
with the usage of metaphor, adequate user feedback and visual
appearance while focusing on the potential user audience and
medical image properties.
III. DISCUSSION
Overall design of the system is present in Fig.3. The
implementation of user interface has been an important aspect
in the development of the FIWeb and careful evaluation has
been conducted to ensure usability. There are no standard in
medical image processing but there has been many
commercial applications such as the ‘photoshop’ and ‘paint
Figure 4. Region of Interest selected on the main picture (left)
and a new picture only consisting of the slected region (right).
The described system to access and diagnose medical
images has several advantages. No software is required on the
viewing site, thus being cost-effective and no installation of
software. This also means that the system can be used outside
the hospital environment and at all time of the day. Also with
the medical imaging, there are the requirements of image
processing to extract useful information that can not be
satisfied by transportable format such as printed/film media
copies but possible with the FIWeb.
This paper can be viewed as archiving one step closer
towards total online medical imaging system and online
imaging in general. It is hoped that the application developed
will provoke more interest in the enormous potential of being
online and the exponential growth of power in computers to
improve the field of medical imaging.
The main reason for the selection of Java as the
programming language is the support of online development
via the Applet software running within the Internet browser.
The extensive image processing libraries available in Java
along with tools for graphical user interface development, ease
of development, and object-oriented programming; use of Java
is very ideal for such system as the FIWeb. There are other
alternatives such as the C++ language but none offered the
similar functionality of the Java when Internet programming
was the criteria. The use of python scripting language over
others such as 'Perl' and 'shell script' for the FIWeb database is
primarily due to the database being object-oriented and filebased which Python scripting languages are best suited.
This research is partially supported by ARC and UGC
grants.
The evaluation of network to be used for the system has
produced positive results. Using the standard 56K dial-up
modem, there are slight delays due to downloading of images
and Applet class files but with Cable modem (tested with
'telstra' and 'optus' cable [5]), the result has been adequate.
With the rapid growth in networking technology and the
introduction of broad bandwidth network, the FIWeb system
becomes very accessible and economical.
V. Acknowledgement
REFERENCE
[1] Najjar, L.J, “Using Color Effectively (or Peacocks Can’t
fly)”,Atlanta, GA: IBM Corporation, 1990
[2] Habib Zaidi, “Medical Imaging: Current status and
future perspective”
[3] Piotr J. Slomka, Edward Elliot, and Albert A. Driedger ,
“Java-Based Remote viewing and processing of Nuclear
Medicine Images: Towards ‘the Imaging Department
Without Walls’”, The journal of Nuclear Medicine, Vol.
41, No.1, pp111-118, January 2000
[4] Piotr J. Slomka, Edward Elliot, and Albert A. Driedger –
Java-based PACS and reporting system for Nuclear
Medicine.
IV. Future work and Conclusion
In this paper, we have presented the potential of archiving
on-line PACS system that is both efficient and functional. We
addresses a broad range of issues relating to online medical
imaging such as the storage and retrieval of medical data in a
well-structured and manageable database to the ability of
online image processing and imaging tools. The creation of
FIWeb is to provide real users with the ability to perform
above conditions through an easy and practical experience.
The area of research of this paper is new and exciting, with
constant improvement in computer performance allowing the
realization of new concepts and technology. Several new
features are planned for development that will make the
system more complete and powerful. The inclusion of
integrated database to the FIWeb, meaning that the query to
the database can be done within the FIWeb system rather than
independently, could enhance performance and usability. With
the strong growth in the 3-D imaging such as the volume
representation, the support for such technique will provide
greater resource to the physicians to perform diagnosis and
study.
[5] Jerold W. Wallis, “Java and Teleradiology (Invited
commentary)”, The journal of Nuclear Medicine, Vol.
41, No.1, pp119-122, January 2000
[6] Prashant Jain, Seth Widoff, and Douglas C. Schmidt,
“The Design and Performance of MedJava”
[7] H.K. Huang, “PACS: Basic Principle and Applications”,
Wiley-Liss, NY, 1999.
[8] Yanxi Liu, Frank Dellaert and William E. Rothfus,
“Classification Driven Semantic Based Medical Image
Indexing and Retrieval.”, 1998
[9] J. Hu, S. Mungee, and D. C. Schmidt, “Principle for
developing and Measuring High-performance Web
Servers over ATM,” in Proceedings of INFOCOM ’98,
March/April 1998.
[10] J.Hu, I. Pyarali, and D. C. Schmidt, “Measuring the
Impact of Event Dispatching and Concurrency Models
on
Web
Server
Performance
Over
High-speed
Networks”, in Proceedings of the 2nd Global Internet
Conference, IEEE, November 1997.
[11] Venkat N. Gudivada and Vijay V. Raghavan, “Picture
Retrieval Systems: A Unified Perspective and Research
Issues”
[12] W. Stallings. Cryptography and Network Security:
Principle and Practice, 2nd edition, Prentice Hall1995,
1999.
[13] Chronaki,
Tsiknakis,
Zabulis,
and
Stathiakis,
Leiscg,
Orphannoudakis,
“A
Katehakis,
Web-based
Environment for Medical Collaboration in a Regional
Healthcare Network”, Elsevier Science B.V, 1997
[14] R. C. Gonzalez, Richard E. Woods. Digital Image
Processing, Addison Wesley, 1993
[15] Walrath K, Campione M. The JFC Swing Tutorial,
Addison-Wesley, 1999.
[16] R. Dumas, A practical guide to usability testing, 2nd
edition, Ablex Publishing Corporation, 1994.