Quantifying Hippocampal Shape Asymmetry in Alzheimer's Disease Using Optimal Shape Correspondences

TL;DR

The paper addresses localization of hippocampal shape asymmetry in Alzheimer's disease beyond volume metrics. It proposes a point distribution model built with ShapeWorks to optimize interhemispheric and population-wide correspondences and quantifies asymmetry per point with $y_i^m = (\\mathbf{d}_i^m)\\cdot \\mathbf{n}_i^m$ and $E_i = (l_i + r_i)/2$, using a joint cost $Q = \\alpha H(\\mathbf{Z}) - \\sum_{i=1}^N H(\\mathbf{X}_i)$ and a linear model with covariates. On the OASIS3 dataset, shape asymmetry reveals fine-grained directional differences not captured by volume alone, with 177 of 512 surface points significant after FDR and a Hotelling $T^2$ test showing strong group separation ($p$-value numerically zero). This framework yields intuitive visualization and potential clinical utility for AD prognosis and can be extended to other brain structures.

Abstract

Hippocampal atrophy in Alzheimer's disease (AD) is asymmetric and spatially inhomogeneous. While extensive work has been done on volume and shape analysis of atrophy of the hippocampus in AD, less attention has been given to hippocampal asymmetry specifically. Previous studies of hippocampal asymmetry are limited to global volume or shape measures, which don't localize shape asymmetry at the point level. In this paper, we propose to quantify localized shape asymmetry by optimizing point correspondences between left and right hippocampi within a subject, while simultaneously favoring a compact statistical shape model of the entire sample. To account for related variables that have impact on AD and healthy subject differences, we build linear models with other confounding factors. Our results on the OASIS3 dataset demonstrate that compared to using volumetric information, shape asymmetry reveals fine-grained, localized differences that indicate the hippocampal regions of most significant shape asymmetry in AD patients.

Figures (3)

• Figure 1: Normalized directional volume asymmetry and undirectional volume asymmetry
• Figure 2: Superimposed left and right mean hippocampus for AD and healthy subjects
• Figure 3: Linear model statistics of diagnosis and statistically significant areas after FDR correction(blue dots)