Program details

David Sarrut - CREATIS lab, medical imaging and AI

This talk will give an overview of the research in cancerology performed at the CREATIS lab, which is primarily focused on medical imaging. We develop innovative machine/deep learning methods for analyzing multimodal images (MRI, US, X-ray, nuclear) with applications in early detection, image quantification, personalized treatment planning, dosimetry, and image reconstruction

Arnaud Bonnafoux - CRCL, Retro-engineering of Gene Regulatory Networks with data exploration tools and machine learning

The challenge of GRN inference is a transdisciplinary team work between biologists, bioinformaticians, and software developers. Biologists should be involved in performing the feature selection critical step prior to the step of retro-engineering with ML. In this presentation, I will present our single-cell data analysis platform TWIST designed for biologists, and our ML algorithm WASABI to infer GRNs that have been experimentally validated in 2 use cases.  

Quentin Filori - CLB, Artificial Intelligence in Clinical Routine at CLB: Current Status, Projects, and Reflections

This presentation provides an overview of AI tools currently implemented in clinical routine at Centre Léon Bérard, as well as ongoing projects and pilot studies. It explores the challenges, opportunities, and early impacts of AI on medical practice for both physicians and patients, drawing on concrete examples and lessons learned from CLB’s experience.

Olivier Cochet-Escartin - ILM, Examples of the use of hand-designed CNNs for biophysical image analysis

In this talk, I will present how our experimental biophysics team leverages machine learning tools to help with experiment analysis. I will focus on applications of convolutional neural networks to the automated treatment of microscopy images in different contexts and end with perspectives for future projects.

Nicolas Alcala - IARC, Combining AI and molecular data to refine cancer classification

Accurate tumor classification remains central to cancer diagnosis, prognosis, and treatment decisions. While traditional histopathological classifications are well-established, genomic studies have uncovered molecular subtypes with significant clinical implications—often presumably invisible to the human eye. However, incorporating high-dimensional molecular data into practical classification systems presents significant challenges. I will present approaches that combine supervised and self-supervised deep learning models, attention mechanisms, and expert pathological annotations to bridge this gap. Using case studies from thoracic tumors, I will show how these models can assess the prognostic relevance of existing classifications and reveal the presumably invisible morphologically differences between molecular subtypes to empower the next generation of cancer classifications.

Hilary Robbins, IARC, Risk prediction and biomarkers in lung cancer screening

Screening by low-dose computed tomography (LDCT) can reduce lung cancer mortality among people with a substantial smoking history. Tools such as risk prediction models and biomarkers can be used to better balance the benefits and harms of screening. This presentation will describe research in the IARC Risk Assessment and Early Detection (RED) team, much of which has been conducted within the Lung Cancer Cohort Consortium (LC3). The LC3 brings together data and biospecimens from 3 million participants from 26 cohorts worldwide (lc3.iarc.who.int).

Nikolina Trendov (M2 Cancer Bioengineering, team Loïc Verlingue, CRCL) Autonomous AI Agents in Oncology and Recent Progress on the PubMed Tool for Clinical Decision Making

This presentation introduces the work of Dr. Verlingue’s group within a multi-agent AI framework for oncology, with a focus on recent progress on the PubMed Tool. This tool enhances literature search to support molecular tumor boards in complex clinical decision-making.

I’m a second-year Master’s student in Cancer Bioengineering, having completed a Bachelor’s in Microbiology and the first year of my Master’s in a Cancer program. During my internship in Dr. Verlingue’s group, I discovered a strong interest in using AI to improve patient care and support innovation in oncology.