Deep learning-based prediction of “brain age” and cognitive decline
Faculty:Sridharan Devarajan (CNS & CSA), Jonas Sundarakumar (CBR)
Description
Age-related neurological disorders, like Alzheimer’s dementia (AD), are increasingly common in the aging population, both in India and across the world.
- In this project we will analyze a large database of brain imaging scans (structural, functional and diffusion MRI) acquired from a large cohort of cognitively aging healthy participants, as well as patients with cognitive decline collected at IISc, Bangalore.
- We will then apply state-of-the-art deep learning approaches, including graph convolutional networks, for predicting brain health (or “brain age”) and the onset of cognitive decline.
- We will also explore more advanced models, like generative diffusion models as well as vision transformers for minimally supervised feature extraction and domain generalization.
- The results will be relevant for identifying important biomarkers of brain aging in a healthy, elderly population, as well as in patients with neurological disorders, like Alzheimer’s disease.

Reference: Gurusamy et al. Diffusion MRI-based structural connectivity robustly predicts”brain-age”. NeurIPS – Workshop on Medical Imaging, 2019.