This lecture introduces models and algorithms for analyzing temporal sequences of images to detect and quantify changes. The main clinical application discussed is the measurement of brain atrophy in Alzheimer's disease, but other diseases are briefly illustrated, notably multiple sclerosis. The algorithms are guided mainly by advanced geometric models (diffeomorphisms parameterized by velocity streams) and adapted statistical models. A biophysical model of atrophy is briefly presented, along with future prospects.
This lecture is complemented by the following two seminars:
- "Neuroimaging in the era of the digital patient", Stéphane Lehéricy (IHU Pitié-Salpêtrière);
- Imaging biomarkers in brain pathologies", Christian Barillot (CNRS, Inserm, Irisa).
