AI & Cognitive Neuroscience


The overall goal of this research line is to develop artificial neuronal network approaches that reproduces brain MR imaging data, representing the next neuroscience revolution (“neurobiologically-plausible artificial intelligence (AI) models”), fundamentally advancing our understanding of human brain function. We cannot develop more advanced models of brain functioning without a more thorough understanding of brain dynamics and networks. However, our understanding of the human brain is limited by biased approaches and limitations of traditional statistics, thus forcing us to rely on artificial machine learning algorithms to provide an alternative perspective. The goal is to dynamically and comprehensively describe cognitive processes as an emergent property of brain networks on healthy subjects and patients. This will then allow characterizing human brain function on a single-subject level for the purpose of personalized medicine. In the individual projects of this research line, we develop and apply advanced AI methods on human fMRI data and develop dynamic generative models of brain connectivity (aka Dynamic Causal Modeling).


  • Deep Image Reconstruction from Human Brain Activity using Capsule Networks (Shawn)
  • Dynamic Effective Connectivity (Sayan)
  • Neuroscience-inspired Deep Net Architectures (Labeeb)


  • Shawn Carere
  • Sayan Nag
  • Labeeb Talukder