Sites Inria

Version française

ARAMIS Research team

Algorithms, models and methods for images and signals of the human brain

Team presentation

Our team is integrated within the Brain and Spine Institute (ICM). It is a joint team with CNRS, Inserm and University Pierre and Marie Curie. The ICM is a recently created neuroscience center based in the Pitié-Salpêtrière hospital in Paris, which is the largest adult hospital in Europe and has a long tradition of neuroscience and neurology. This rich environment allows direct and fruitful collaborations with medical and biological teams, which is crucial for the design of meaningful methods and for the translation to clinical applications.

The general objective of the team is to design new approaches to study the structural and functional aspects of the human brain, based on neuroimaging and electrophysiological data. For the first aspect, we aim to model brain structure from magnetic resonance imaging (MRI) data. To that purpose, we design geometrical, morphometric and statistical approaches to study the variability of anatomical structures and their alterations in neurological disorders. For functional aspects, the challenge is to better understand how different brain regions cooperate within complex networks. To that purpose, we design methods to extract, analyze and model those networks from EEG (electroencephalography) and MEG (magnetoencephalography) data. Finally, we design methods to integrate multimodal data, in particular neuroimaging and genomics, in order to study the influence of genetic factors on anatomical shapes and their alterations in neurological disorders.

We develop various clinical applications of our research, in particular in neurodegenerative disorders (Alzheimer's disease and other dementias), epilepsy, neurodevelopmental disorders and to design brain-computer interfaces for rehabilitation.

Research themes

Main research themes:
  1. Modeling brain structure from anatomical and diffusion MRI
  2. Modeling dynamical brain networks from EEG/MEG
  3. Integrating multimodal data (neuroimaging, genomics, clinical data)
Key methodological domains:
  • morphometry, statistical shape analysis, diffeomorphic registration
  • image segmentation
  • complex networks theory
  • graph analysis
  • machine learning
  • longitudinal models
  • multi-site MRI harmonization, quality control and analysis
  • ultra-high field MRI (7 Tesla MRI)

Keywords: Neuroimaging Image analysis Signal processing Statistical models Neurological disorders