80-Prime
The project is part of a global interdisciplinary context concerning brain molecular imaging. The objective is to develop statistical methodologies for the classification of textures using a Bayesian approach which combines random field modeling with Deep Convolutional Neural Networks. These methodologies are designed to process data from molecular imaging by Positron Emission Tomography (PET) and Mono-Photon Emission Tomography (SPECT) in order to improve the management of neurodegenerative brain diseases, by exploiting spatial, temporal and multi-parametric (multi-target) information from these imaging modalities.
- Funded by the ANR.
- Collaboration with Eric Guedj.
- Supervision of the PhD thesis of Farideh Bazangani.
Frédéric J.P. Richard
Professor of Applied Mathematics
Professor of applied mathematics whose research interests include image processing, spatial statistics and data science.