Learning Cortical Anomaly through Masked Encoding for Unsupervised Heterogeneity Mapping

TL;DR

We address the challenge of detecting heterogeneous mental disorders from brain imaging by proposing CAM, a self-supervised framework that learns cortical representations from surface features via masked image modeling. CAM employs an iterative masked anomaly-detection strategy to quantify deviations from learned representations, enabling unsupervised detection of psychotic-spectrum disorders. In experiments on left-hemisphere cortical features, CAM outperforms baseline statistical methods and traditional DL models, with notable AUCs up to 0.769 for Schizophreniform and identification of ROIs such as Pars Triangularis and frontal regions as potential biomarkers. The approach is scalable, label-free, and extendable to multi-site and multimodal data.

Abstract

The detection of heterogeneous mental disorders based on brain readouts remains challenging due to the complexity of symptoms and the absence of reliable biomarkers. This paper introduces CAM (Cortical Anomaly Detection through Masked Image Modeling), a novel self-supervised framework designed for the unsupervised detection of complex brain disorders using cortical surface features. We employ this framework for the detection of individuals on the psychotic spectrum and demonstrate its capabilities compared to state-of-the-art methods, achieving an AUC of 0.696 for Schizoaffective and 0.769 for Schizophreniform, without the need for any labels. Furthermore, the analysis of atypical cortical regions, including Pars Triangularis and several frontal areas often implicated in schizophrenia, provides further confidence in our approach. Altogether, we demonstrate a scalable approach for anomaly detection of complex brain disorders based on cortical abnormalities. The code will be made available at https://github.com/chadHGY/CAM.

Figures (1)

• Figure 1: Our proposed Cortical Anomaly detection through Masked image modeling (CAM) framework.