International Transfer of Stochastic Cortical Self-Reconstruction

2026-08-07

AI Summary

This study addresses the limitations of traditional cortical atrophy modeling—coarse regional granularity, reliance on covariates, and poor generalizability—by leveraging the Stochastic Cortical Self-Reconstruction (SCSR) framework. It presents the first validation of SCSR’s robust cross-population transferability, successfully generalizing from the UK Biobank cohort to a Chinese population. The approach employs vertex-level personalized healthy references, integrating a spherical U-Net (SUNet) with a multilayer perceptron for reconstruction, and incorporates fine-tuning strategies to enhance performance. The fine-tuned SUNet achieves state-of-the-art results with an average pairwise AUC of 0.848 and exhibits low reconstruction error across the full lifespan, significantly improving the detection of Alzheimer’s disease–related cortical atrophy in diverse populations.

Abstract

Stochastic cortical self-reconstruction (SCSR) enables personalized mapping of gray matter atrophy, a hallmark of neurodegenerative disorders such as Alzheimer's disease (AD), onto high-resolution cortical surfaces. Unlike conventional normative modeling approaches, which typically operate at a coarse regional level and remain inherently constrained by the covariates included during training, SCSR estimates an individualized healthy reference directly from the observed cortical thickness at the vertex level. This allows the detection of subtle, subject-specific deviations from healthy cortical shape. In this work, we investigate the generalization and transferability of SCSR, originally trained on UK Biobank (UKB) data, to an independent Chinese population dataset. Specifically, we evaluate the ability of SCSR-derived Z-scores to discriminate between healthy scans, individuals with mild cognitive impairment (MCI), and patients with AD, while also assessing model robustness across the lifespan. We compare four training strategies: direct application of the UKB-trained model, fine-tuning on Chinese data, training from scratch, and joint training on UKB and Chinese cohorts. As reconstruction backbones, we consider both a multilayer perceptron (MLP) and a Spherical UNet (SUNet). Our results demonstrate that SCSR provides robust detection of cortical atrophy in the Chinese population across all evaluated models. The highest discriminative performance was achieved by the fine-tuned SUNet model (average pairwise AUC = 0.848), followed closely by the UKB-trained SUNet. Moreover, reconstruction errors remained low across the lifespan, even when the training population exhibited a substantially narrower age distribution, indicating strong cross-population transferability.

Problem

• stochastic cortical self-reconstruction
• cross-population transferability
• cortical atrophy
• Alzheimer's disease
• generalization

Innovation

• stochastic cortical self-reconstruction
• cross-population transferability
• vertex-level cortical modeling
• Spherical UNet
• personalized normative modeling