Artificial intelligence for the evaluation of profiles, trajectories and management of vulnerability in health: COVID-19, frailty of the elderly and cancer
Development of an AI algorithm for prostate MRI interpretation
Elaboration of a deformable anatomical atlas for the diagnosis of prostate cancer in MRI. 1st step: automatic zonal segmentation of a prostate in MRI in order to detect/contour the organ and the 2 main zones, to be able to localize and characterize the lesions.
© François Marin/AP-HP
The algorithm used is a neural network, with a manually segmented training and test base. 2nd step: detection and characterization of lesions (using a PI-RADS score).
Inria, Université Côté d'Azur, AP-HP