Research structure

The AI-RACLES Chair in Artificial Intelligence, created in 2020 and co-directed by Etienne Audureau (AP-HP), Thomas Guyet (INRIA), Laurent Le Brusquet and Arthur Tenenhaus (CentraleSupélec), aims at exploiting the massive data of the AP-HP's Health Data Warehouse (HDW) in order to carry out research focused on the exploration of the concept of vulnerability in health, whether related to ageing or to pathologies such as cancer or COVID-19.

Around this applicative research theme, the AI-RACLES Chair relies on the support of data scientists and the supervision of doctoral and post-doctoral students to develop a research program in three main areas aimed at meeting the challenges related to the exploitation of heterogeneous and longitudinal DHS data:

- Axis 1 - Integration of heterogeneous data at a given time and/or from longitudinal follow-up, including clinical, biological, textual data (hospitalization reports, imaging, anatomopathology, etc.)

- Axis 2 - Identification of frailty phenotypes and care trajectories involving the conduct of unsupervised analyses (representation of the patient and identification of typical profiles and trajectories)

- Axis 3 - Construction and integration of predictive tools useful for clinical practice involving the conduct of supervised analyses based in particular on machine learning approaches