Improving Sensitivity of Early Detection of AD via Multidimensional Analysis of Longitudinal MR Scans

Faisal Beg

Abstract

We propose to use recent Computational Anatomy (CA) algorithms from our group to develop novel multidimensional longitudinal biomarker for early detection of neuroanatomical change due to AD vis--vis normal aging. The overarching hypotheses for our proposal are that a) multidimensional biomarkers containing combined neuroanatomical measures from several structures at single time-point (baseline) and b) longitudinal multidimensional biomarkers additionally combining rate of change in neuroanatomical measures over time (follow-up scan compared to baseline) will offer superior sensitivity for early detection of AD than using neuroanatomical measures of individual structures separately at baseline alone.