RECA-PD: A Robust Explainable Cross-Attention Method for Speech-based Parkinson's Disease Classification

TL;DR

RECA-PD introduces a robust, explainable cross-attention model for speech-based Parkinson's Disease detection, achieving high accuracy while providing clinically meaningful explanations and addressing challenges in long speech recordings.

Contribution

It presents a novel explainable architecture combining interpretable features with self-supervised representations for PD detection from speech.

Findings

Matches state-of-the-art accuracy in PD detection

Provides more consistent and clinically meaningful explanations

Segmenting long recordings improves performance in certain speech tasks

Abstract

Parkinson's Disease (PD) affects over 10 million people globally, with speech impairments often preceding motor symptoms by years, making speech a valuable modality for early, non-invasive detection. While recent deep-learning models achieve high accuracy, they typically lack the explainability required for clinical use. To address this, we propose RECA-PD, a novel, robust, and explainable cross-attention architecture that combines interpretable speech features with self-supervised representations. RECA-PD matches state-of-the-art performance in Speech-based PD detection while providing explanations that are more consistent and more clinically meaningful. Additionally, we demonstrate that performance degradation in certain speech tasks (e.g., monologue) can be mitigated by segmenting long recordings. Our findings indicate that performance and explainability are not necessarily mutually exclusive. Future work will enhance the usability of explanations for non-experts and explore severity estimation to increase the real-world clinical relevance.