1- Scientific context and project objectives

For certain cancer treatments using ionising radiation, the treatment planning software used can become inaccurate. New strategies therefore need to be set up to improve cancer treatment by radiotherapy so as to radiate the tumour as accurately as possible and limit the radiation of healthy tissue and organs at risk. To meet these requirements, the accuracy of Monte Carlo simulations is used to model the ballistics of radiotherapy and curietherapy treatments and to study the dose deposits incumbent upon this radiation.

For 3 years now, the Jean Perrin Cancer Centre and Life Sciences Computing Platform (PCSV) team at the Corpuscular Physics Laboratory have been studying use of the GATE simulation platform for the dosimetry and acceleration of treatment computation on the Auvergrid and EGEE grid. In order to use the Monte Carlo GATE simulations for patient treatments, the medical images (generally scans) produced must be integrated in the simulations to visualise the tumour and at-risk organs and to carry out the necessary contouring for planning treatments.
The project objective is therefore to develop an information system for using medical images in the modelling of the dose deposit by Monte Carlo simulations on a grid.

2- Project description

Two major aspects are to be developed for this project, taking account of the PCSV team’s developments over the last three years in partnership with the radiotherapy department at the Jean Perrin Centre:

• 1°) Optimised distribution of Monte Carlo grid computations. This involves:
  • a) Optimum management of grid simulations to make significant computing time-savings.
  • b) Transparent submission maintaining the confidentiality of medical data from GATE simulations for medical personnel and research stakeholders.
• 2°) Management, storage and secure access to medical images. This involves:
  • a) Development of an interface between the PACS system and the computing grid
  • b) Secure, optimum management of medical images on the grid for Monte Carlo GATE simulations

3- Public or controlled use

The Monte Carlo simulations are intended to be used by carers (radiotherapists, doctors and medical physicians) who want to plan a radiotherapy or curietherapy treatment demanding greater accuracy.
The software developments providing such a service use open source technologies. The medical data will be managed internally by the hospital facility and be accessible to authorised users.
Access to the computing grid is closely controlled by the use of certificates, passwords and usernames as well as an encryption of medical images when these are positioned on grid storage elements.

4- Expected results

Through various partnerships on:

• The development of the Monte Carlo GATE simulation platform with OpenGATE (http://www.opengatecollaboration.org).
• The distribution of the platform on computing grid at regional level for the AuverGrid project and at European level for the EGEE project.
• The interfacing of the PACS system of the Jean Perrin CLCC with its IT department and the assistance of the system provider.

The project must enable the optimised distribution of Monte Carlo grid simulations. Once optimum use of the computing and storage facilities has been achieved, a Web portal to the grid will give CLCC staff user-friendly and transparent access to resources for planning treatments with Monte Carlo. Medical images are managed in respect of data confidentiality through direct access to the PACS connected to the computing grid. Medical images no longer correlated to their confidential contents will then be used by Monte Carlo grid computations through an additional encryption of images and authorised access on storage elements.

Thanks to these developments, the project will significantly improve certain radiotherapy treatments using cutting-edge techniques or complex radiation.

LifeGrid, the regional information system