VoIP Service and Marketing using Odoo and Asterisk PBX
EuroPACS 2006
1. Faculty of Medicine
University of Crete,
Greece
A RIS/PACS system
with integrated advanced image processing technology,
which is extensible by end-users (in Java or Matlab)
Kostas P. Karolemeas1
, Thomas G. Maris2
, A. Karantanas3
, N. Gourtsoyiannis3
1
Evorad – Medical Information Systems, Athens, Greece
2
Medical Physics Department, Faculty of Medicine, University of Crete, Heraklion, Greece
3
Radiology Department, Faculty of Medicine, University of Crete, Heraklion, Greece
Kostas P. Karolemeas
Research Fellow, University Hospital of Heraklion, Greece
Chief Technology Officer, Evorad
EuroPACS 2006
2. Faculty of Medicine
University of Crete,
Greece Case Study
End-user extensible image processing technology
provided by:
applied to:
RIS/PACS & Research PACS
www.evorad.com
Faculty of Medicine
University of Crete,
Greece
University
Hospital
of Heraklion
www.med.uoc.gr www.pagni.gr
3. Faculty of Medicine
University of Crete,
Greece Previous Status
Time consuming image and data transferring between
independent workstations
In-house developed software required two many steps to
complete
Lack of algorithm documentation and limited
customization in commercial post-processing software
Significantly complex image processing for radiologists to
apply on a regular basis
Remote researchers needed extra software apart from
exchanging e-mails and CDs with images
4. Faculty of Medicine
University of Crete,
Greece Requirements
Advanced medical image processing:
rapidly implemented or customized by end-users
integrated with regular RIS/PACS work-flow for use by
radiologists and technologists
with batch-processing capabilities for clinical research
integrated with external tools for advanced mathematical
calculations
with data extraction capabilities for further processing
(e.g. statistics)
with support for geographically dispersed teams
with results' attachment to exams
5. Faculty of Medicine
University of Crete,
Greece Evorad RIS/PACS
Visual examination
scheduling
Full examination work-flow
EHR with images and HL7
integration
Voice recording and
speech recognition for
reports
Integrated report editor with
templates
Extensive reporting
Full functionality available
through the Internet
6. Faculty of Medicine
University of Crete,
Greece Evorad RIS/PACS
Full DICOM integration
Extended image manipulation
Imports images from other
exams for comparison
Extensible image processing
technology
Extensible image
importing/exporting
technology
CD/DVD viewer with full
image processing
Full functionality available
through the Internet
7. Faculty of Medicine
University of Crete,
Greece Evorad Research PACS
Image processing is integrated with RIS/PACS exam
work-flow
Researchers can test and update their image processing
algorithms from within the PACS system and then publish
them instantly for use by all authorized local and remote
users (through the Internet).
Researchers can develop their image processing
techniques (called “image processors”) in Java, Matlab or
both
Remote users can access through the Internet the full
functionality available to local users.
Operating system independent (supports Windows, Linux,
Solaris)
8. Faculty of Medicine
University of Crete,
Greece Image Processors (Java)
“Image Processors”:
Written in Java
based on templates
Are dynamically
updated and
published to all
users
Can have
parameters
Can be chained
Have full access to
RIS/PACS database
Can export / import
information / images
Can be published in
image CDs/DVDs“Image processors” can define parameters
to allow the user to configure their
execution
9. Faculty of Medicine
University of Crete,
Greece Image Processors (Matlab)
A Matlab session is started from within PACS
and image data can be exchanged transparently
without external files
Transparent PACS
integration with
Matlab.
Export images /
image series as 2D /
3D arrays to Matlab
Process data in
Matlab
Import resulting 2D /
3D arrays as
images / image
series from Matlab
Import Figures from
Matlab
Execute Matlab
programs as “Image
Processors”
10. Faculty of Medicine
University of Crete,
Greece
Applying Image Processors
The results of “Image processor” application can
be attached to the examination and become
available to other users
“Image Processor”
application:
Images and other
data can be
exported / imported
using customizable
algorithms
1. A set of
images/image series
is selected
2. The required
processor is picked
from a list
3. Any parameters are
set
4. The resulting images
are added to the
current window
11. Faculty of Medicine
University of Crete,
Greece Batch Processing
A provided “image exporter” accumulates user-specified
information (ROI, exam fields) from multiple images to
specified spreadsheets.
A complex combination of actions such as:
creating new series from existing images
applying “image processors”
exporting / importing images and other information
can be recorded, saved under a user-defined name for
future reference, and applied to a set of examinations.
This can accelerate repetitive procedures especially in
clinical research contexts.
12. Faculty of Medicine
University of Crete,
Greece New Status
Streamlining Image Processing
In-house developed image processing algorithms have
been migrated to “image processors”
Integration with Matlab accelerates development work for
new techniques
All techniques are developed and applied to regular exams
even by radiologists within the exams' work-flow
“image processor” parameterization capabilities foster the
development of versatile algorithms
Supporting Collaboration
Remote researchers can access examinations and apply
“image processors” through the Internet with all
functionality available to local users
13. Faculty of Medicine
University of Crete,
Greece New Status (Cont.)
Accelerating Result Extraction
Manual image processing and data extraction procedures
performed repetitively for clinical research are been
replaced by automated sequences of system actions
Significant time reductions between image acquisition
processing results publication
All generated information is now attached to examinations
and "projects"
Secondary Research PACS
Parallel and integrated with central RIS/PACS (could be
provided by any vendor)
Stores special examinations including phantom studies
Allows testing of new image processing techniques before
they are validated to be used by radiologists
14. Faculty of Medicine
University of Crete,
Greece Conclusion
Since the implementation of Evorad RIS/PACS (central)
and Evorad Research PACS:
Researchers can develop and test new image processing
techniques at a fraction of the time required before.
The incorporation of generated images, measurements
and other information in publications has been
significantly accelerated.
New techniques are validated faster and then become
immediately available for application by radiologists on
regular examinations.
Geographically dispersed team members reduced
travelling, eliminated image exchange, and increased
their productivity.
15. Faculty of Medicine
University of Crete,
Greece
Thank you for your attention!
For further information, contact us:
http://www.evorad.com
kostas.karolemeas@evorad.com
http://www.med.uoc.gr
tmaris@med.uoc.gr