Deep learning identifies brain structures that predict cognition and explain heterogeneity in cognitive aging

Krishnakant V. Saboo, Chang Hu, Yogatheesan Varatharajah, Scott A. Przybelski, Robert I. Reid, Christopher G. Schwarz, Jonathan Graff-Radford, David S. Knopman, Mary M. Machulda, Michelle M. Mielke, Ronald C. Petersen, Paul M. Arnold, Gregory A. Worrell, David T. Jones, Clifford R. Jack, Ravishankar K. Iyer, Prashanthi Vemuri

Research output: Contribution to journalArticlepeer-review


Specific brain structures (gray matter regions and white matter tracts) play a dominant role in determining cognitive decline and explain the heterogeneity in cognitive aging. Identification of these structures is crucial for screening of older adults at risk of cognitive decline. Using deep learning models augmented with a model-interpretation technique on data from 1432 Mayo Clinic Study of Aging participants, we identified a subset of brain structures that were most predictive of individualized cognitive trajectories and indicative of cognitively resilient vs. vulnerable individuals. Specifically, these structures explained why some participants were resilient to the deleterious effects of elevated brain amyloid and poor vascular health. Of these, medial temporal lobe and fornix, reflective of age and pathology-related degeneration, and corpus callosum, reflective of inter-hemispheric disconnection, accounted for 60% of the heterogeneity explained by the most predictive structures. Our results are valuable for identifying cognitively vulnerable individuals and for developing interventions for cognitive decline.

Original languageEnglish (US)
Article number119020
StatePublished - May 1 2022


  • Brain reserve
  • Cognitive aging
  • Cognitive heterogeneity
  • Deep learning

ASJC Scopus subject areas

  • Neurology
  • Cognitive Neuroscience


Dive into the research topics of 'Deep learning identifies brain structures that predict cognition and explain heterogeneity in cognitive aging'. Together they form a unique fingerprint.

Cite this