Approved Research
Holistic bigdata lifespan modeling of brain trajectory for automatic diagnostic of neurodegenerative diseases based on advanced deep learning methods
Approved Research ID: 80509
Approval date: January 12th 2022
Lay summary
This project aims at proposing new lifespan models of brain trajectory for automatic diagnostic of neurodegenerative diseases based on artificial intelligence methods.
First, we will release the most complet representation of cerebral anatomy from brain structures to substructures. Second, we will design the next generation of segmentation methods able to analyse the whole brain anatomy at different anatomical levels using high-resolution 3D Magnetic Resonance Images. Finally, we will propose novel statistical models able to perform global analysis of brain changes over the lifespan. Finally, the developed tools will be applied on large-scale existing databases including healthy subjects, patients with Alzheimer's disease, multiple sclerosis and Parkinson's disease, in order to produce knowledge on healthy and diseased brains. Divergence between normal aging and pathological models will be used to develop new artificial intelligence based diagnostic tools.
Our project will use multimodal brain Magnetic Resonance Images, demographics and clinical scores to segmentation all the brain structure and build lifespan models.
The project is organised in 3 milestones. The first milestone will be dedicated to the fusion of existing segmentation protocols to propose the most complet representation of the brain anatomy (1 year). The second milestone will be dedicated to the development of novel tools able to segmentation all the structural components of the cerebral anatomy (1 year). The last one will be dedicated to the development of lifespan model (1 year). The duration of the project is expected to be 3 years, with potential renew.
Thanks to the proposed framework, brain changes over the lifespan and progression of neurological pathologies will be studied at a level of detail never reached before. These contributions will produce novel knowledge on cerebral organisation and new biomarkers for early diagnostics of neurodegenerative diseases.