Abstract #2641

Biological brain age prediction using structural MRI: Insights from dimensionality reduction techniques

Arna Ghosh^1,2, Alba Xifra-Porxas^2,3, Georgios D. Mitsis⁴, and Marie-Hélène Boudrias^2,5

¹Integrated Program in Neuroscience, McGill University, Montréal, QC, Canada, ²Montréal Center for Interdisciplinary Research in Rehabilitation (CRIR), Montréal, QC, Canada, ³Graduate Program in Biological and Biomedical Engineering, McGill University, Montréal, QC, Canada, ⁴Department of Bioengineering, McGill University, Montréal, QC, Canada, ⁵School of Physical and Occupational Therapy, McGill University, Montréal, QC, Canada

The human brain changes with age and these age-related changes have been used as biomarkers for several brain-related disorders. Therefore, being able to accurately predict the biological age of the brain from T1-weighted MR images yields significant potential for clinical applications. The present study evaluates regression models coupled with dimensionality reduction techniques for biological brain age prediction and concludes that Canonical Correlation Analysis (CCA) enhances prediction performance of Gaussian Process Regression (GPR) models. The proposed analysis also reveals brain areas that are strongly anti-correlated with age, in agreement with previous aging studies.

This abstract and the presentation materials are available to members only; a login is required.

Join Here