Hierarchical Multimodal Sleep Representation Learning for Cross-Site Cognitive Impairment Prediction

Xinqi Bao
KTH Royal Institute of Technology


Abstract

Cognitive impairment progressively limits independent living, yet scalable methods for identifying individuals at future risk remain limited. In the 2026 George B. Moody PhysioNet Challenge, our team (Matcha) developed a hierarchical framework for predicting cognitive impairment from overnight polysomnography (PSG). Modality-specific convolutional stems learned a 192-dimensional representation for every 30-s epoch from human and automated sleep annotations together with cross-view self-supervision. Local and full-night Transformers then aggregated the epoch sequence into a patient-level prediction, while a low-capacity record-wise residual represented whole-night sleep summaries and outcome-observation timing. In the three-site leave-one-site-out (LOSO) validation, it achieved a macro age-conditioned AUROC of 0.804. On the official phase hidden validation set, the best Small- and Large-track entries achieved 0.847 and 0.844, respectively, ranking 1st out of 513 entries.