Predicting cognitive decline in a low-dimensional representation of brain morphology

Abstract

Identifying early signs of neurodegeneration due to Alzheimer's disease (AD) is a necessary first step towards preventing cognitive decline. Individual cortical thickness measures, available after processing anatomical magnetic resonance imaging (MRI), are sensitive markers of neurodegeneration. However, normal aging cortical decline and high inter-individual variability complicate the comparison and statistical determination of the impact of AD-related neurodegeneration on trajectories. In this paper, we computed trajectories of thickness changes in a 2D representation of a 62-dimensional manifold of individual cortical thickness measures. To compute this representation, we used a novel, nonlinear dimension reduction algorithm called Uniform Manifold Approximation and Projection (UMAP). We first trained a UMAP embedding on cortical thickness measurements of 6,237 cognitively healthy participants (3,556 women) aged 18 to 100 years old. Once transformed and rotated, its principal axis was shown to be positively associated (r= 0.65) with participants' age. We then projected in this transformed UMAP space, data from longitudinal MRIs of 537 mild cognitively impaired (MCI) subjects and 340 AD subjects from the Alzheimer's Disease Neuroimaging Initiative database. Each participant had multiple visits (n≥ 2), one year apart. Once in the transformed UMAP space, the data was clustered using ak-means variant. Average trajectories between clusters were shown to be significantly different between MCI and AD subjects. Moreover, we showed that some clusters and trajectories between clusters were more prone to host AD subjects. This study was able to differentiate AD and MCI subjects based on their cluster trajectory in a 2D space with a AUC of 0.72 over 2,000 iterations.