Training-Free Cross-Lingual Dysarthria Severity Assessment via Phonological Subspace Analysis in Self-Supervised Speech Representations

TL;DR

This paper introduces a training-free, cross-lingual method for assessing dysarthria severity using phonological subspace analysis in self-supervised speech representations, applicable across multiple languages without labeled pathological data.

Contribution

The authors propose a novel, training-free approach that quantifies dysarthria severity through phonological feature degradation in pretrained speech models, eliminating the need for supervised training.

Findings

All five consonant d-prime features significantly correlate with clinical severity.

The method generalizes across 10 corpora, 5 languages, and 3 etiologies.

Features distinguish controls from severely dysarthric speakers with high significance.

Abstract

Dysarthric speech severity assessment typically requires trained clinicians or supervised models built from labelled pathological speech, limiting scalability across languages and clinical settings. We present a training-free method that quantifies dysarthria severity by measuring degradation in phonological feature subspaces within frozen HuBERT representations. No supervised severity model is trained; feature directions are estimated from healthy control speech using a pretrained forced aligner. For each speaker, we extract phone-level embeddings via Montreal Forced Aligner, compute d-prime scores along phonological contrast directions (nasality, voicing, stridency, sonorance, manner, and four vowel features) derived exclusively from healthy controls, and construct a 12-dimensional phonological profile.Evaluating 890 speakers across 10 corpora, 5 languages (English, Spanish, Dutch, Mandarin, French), and 3 primary aetiologies (Parkinson's disease, cerebral palsy, ALS), we find that all five consonant d-prime features correlate significantly with clinical severity (random-effects meta-analysis rho = -0.50 to -0.56, p < 2e-4; pooled Spearman rho = -0.47 to -0.55 with bootstrap 95% CIs not crossing zero). The effect replicates within individual corpora, survives FDR correction, and remains robust to leave-one-corpus-out removal and alignment quality controls. Nasality d-prime decreases monotonically from control to severe in 6 of 7 severity-graded corpora. Mann-Whitney U tests confirm that all 12 features distinguish controls from severely dysarthric speakers (p < 0.001).The method requires no dysarthric training data and applies to any language with an existing MFA acoustic model (currently 29 languages). We release the full pipeline and phone feature configurations for six languages.