Label-Free Parkinson's Disease Screening from Face and Voice through Mechanistic Interpretability
2026-08-10 • Machine Learning
Machine Learning
AI summaryⓘ
The authors developed a way to screen for Parkinson's disease (PD) without needing labeled patient data, using only pre-trained models for face expressions and voice. They created synthetic voice changes to help detect PD-related speech issues and used an anomaly detection method for facial expressions. Their method works well for voice data but less so for faces, and combining both gives decent accuracy for ruling out PD. They note that the face results might be optimistic and need more testing on other data.
Parkinson's diseasemachine learningVision TransformerHuBERTcontrastive activation additionanomaly detectioncosine similarityAUROCsynthetic dysarthriadata privacy
Authors
Jiaheng Su, Yu Sun
Abstract
Parkinson's disease (PD) is the second most common neurodegenerative disorder. Typical machine learning screening methods require PD labels, but the available data is limited by privacy concerns and the need for expert annotation. We propose a label-free face-plus-voice PD screen built entirely on frozen pretrained encoders--a face-expression Vision Transformer and HuBERT--in which no PD label touches any fit; the reference is training controls only. The voice modality uses a synthetic-dysarthria contrastive activation addition (CAA) direction built from time-stretch and breathy degradation of healthy speech; the face modality uses a k-nearest-neighbor anomaly score to the control embedding cluster. We introduce the alignment principle, a post-hoc analysis showing that a synthetic-degradation CAA detector works when the cosine similarity between the synthetic and real disease directions exceeds zero. Measured on the YouTubePD benchmark, this cosine is +0.37 for voice (CAA works, AUROC 0.765) and -0.48 for face (CAA fails; anomaly succeeds, AUROC 0.751). Equal-weight late fusion reaches AUROC 0.802 (95% CI [0.70,0.89]) with NPV 0.95, supporting a rule-out triage interpretation. An overfitting audit shows the voice detector transfers cleanly, while the face-side--and thus fused--AUROC is potentially optimistic pending external validation.