Tiers Lieu : ESPACE
Lead partner: Centre Léon Bérard (Lyon)
Main consortium members: Center Oscar Lambret (Lille), Centre Francois Baclesse (Caen), Laboratoire CREATIS - CNRS (Lyon)
ESPACE aims to accelerate the development and deployment of 5P medicine (preventive, personalized, predictive, participatory and evidence-based), in oncology, thanks to Artificial Intelligence (AI).
ESPACE - Patient-centered experiments for the advancement of artificial intelligence in oncology
90 000
patients followed every year, only in oncology
.800
publications per year
30
experts mobilized to assess and support projects
The Tiers Lieu offer
- Facilitated access to medical expertise and care organization : ESPACE enables ideas to be confronted and needs to be refined to better design solutions. This strategy also helps to identify uncovered needs, encouraging the launch of new projects.
- Access to high-quality healthcare data to evaluate proofs of concept and make solutions under development more reliable.
- Access to rigorous evaluation by recognized experts (regulatory, ethical, economic and methodological) and to ensure smooth integration of AI "modules" into practice and the healthcare system
Ongoing experimental projects
Project: TheraSPACE
Supplier: Therapanacea
Solution: MR-BOX enables radiotherapy treatment planning using MRI instead of CT to improve accuracy. The tool generates a synthetic CT by artificial intelligence from MRI
Need : Modern radiotherapy faces two challenges: precision targeting of volumes to be treated and daily adaptation of the therapeutic plan. MRI offers better visualization, enabling more precise and less toxic planning, with adaptive treatments. However, an additional CT scanner is required for dose calculation. MR-BOX enables treatment planning from a single MRI image, creating a synthetic CT via AI.
Third-Party Contribution: Two prospective studies will be initiated over two years, each involving 150. One study will focus on head and neck cancers at CLB, the other on localized prostate tumors at COL. Outcomes evaluated will include clinical consensus, workflow improvements and socio-economic factors.
Project: SoftCHEMO
Vendor : OSPI
Solution : The SoftChemo project is based on an innovative language model that predicts the frequent medical causes of hospitalization in oncology. This model analyzes medical reports and includes a user interface.
Need : Around 60% of patients undergoing chemotherapy for cancer experience a severe medical event within 6 months, representing almost 220,000 patients nationwide. ChemoSoft aims to improve the monitoring and early management of these events between town and hospital using AI, thereby reducing hospitalization rates through personalized prevention.
Third-Party Contribution : An in-depth prospective evaluation of the tool will be carried out on two complementary day hospitals, enabling a POC to be tested and the tool to be evaluated with a view to CE marking.