Transcript telemedicine_poster

Grid based telemedicine application
Enabling Grids for E-sciencE
Grid based telemedicine application for Monte Carlo dosimetric studies
Matteo Diarena, Arnaud Fessy, Vincent Breton, Yannick Legré, Deng Ning, Simon Nowak, Lydia Maigne
Laboratoire de Physique Corpusculaire, 24 avenue des Landais, 63177 Aubière cedex
Introduction
Hospital
SE
SE
SE
SE
Job output
Physicians and medical physicists through the GridSphere based portal are able to manage patient
information, medical images and simulation jobs. Medical information and images metadata are stored in
the local AMGA server while the encrypted images are stored in Grid SE to be run in Monte Carlo
simulations for dosimetric calculations. All the interactions between users and the Grid are managed by
the telemedicine Web Services deployed in the hospital.
AMGA
Server
Monte Carlo Simulations
GridSphere
Portal
Metadata
Images (AES)
Platform structure
SOAP
Web Services
Nowadays the need for communication facilities between
physicians located in different hospitals far from each other is a
growing request. Remote consulting is even more important for
places where specialist consulting are expensive and difficult to
achieve (for example places with poor transportation facilities).
The aim of this project is to provide physicians an easy to use
medical platform that allow to easily exchange medical information
and, using Grid resources, give them the possibility to run complex
Monte Carlo simulations used in cancer diagnostic.
CE
Medical data sharing
The telemedicine infrastructure allows users to exchange medical data from different hospitals. The information exchange takes place
between web services deployed in different locations and is completely transparent to end users. Information are transferred
between different locations only on request using SOAP messages over an SSL secured channel.
Requirements
Gridsphere Portal
SOAP
Web Service
AMGA API
AMGA
Server
gLite UI
Hospital B
SOAP (SSL)
Hospital A
Performance and reliability:
- Have to work in places with poor communication
infrastructure (i.e. Burkina Faso & China hospitals)
- Monte Carlo simulation is a cpu-intensive task that
can have good performance improvements using grid
gLite API
gLite API
gLite UI
SOAP
Web Service
Gridsphere Portal
AMGA API
AMGA
Server
Statistics and Results
Simulation job’s execution time in a local
machine compared to the Grid
Security:
- Communications encrypted using SSL
- Medical images stored encrypted on Grid
- AMGA server is local to the hospital
Usability:
- Physicians don’t want to change their way of working
- Need to provide an easy to use interface
- Need to hide grid complexity to end users
Conclusions
Results of dosimetric simulation are stored in the
AMGA server and their visualization is available
from the web portal
Grid technology allows collaboration and resource sharing
between users all over the world. Our telemedicine platform
exploits this big amount of resources to offer physicians a
set of tools to easily share medical data and resources.
This resource sharing allows doctors to run in a reasonable
time complex simulation software to diagnose cancer and
help defeating this dreadful disease.