A Hierarchical Feature Engineering Framework for Automated Classification of Phonotraumatic and Non-Phonotraumatic Vocal Hyperfunction

2026-06-04

AI Summary

This study addresses the critical lack of effective biomarkers for differentiating phonotraumatic vocal hyperfunction (PVH) from non-phonotraumatic vocal hyperfunction (NPVH). To this end, the authors propose a hierarchical feature engineering framework that, for the first time, systematically incorporates source–filter coupling features to capture the interaction between the glottal source and vocal tract. The framework integrates static, dynamic, and ratio-based acoustic features and leverages high-dimensional machine learning models for automated classification. Evaluated on the NeckVibe Challenge dataset, the approach significantly improves diagnostic performance, achieving area under the curve (AUC) scores of 0.891 for PVH and 0.728 for NPVH, thereby demonstrating the pivotal role of source–filter coupling features in enhancing NPVH identification.

Abstract

Ambulatory neck-surface acceleration enables non-invasive monitoring of vocal hyperfunction, yet robust biomarkers for its subtypes remain limited. This study investigates the NeckVibe Challenge dataset to distinguish phonotraumatic (PVH) and non-phonotraumatic (NPVH) from healthy controls. We propose a hierarchical feature engineering framework comprising: (i) static, (ii) dynamic, (iii) ratio-based, (iv) coupling features capturing source filter interactions. While univariate statistical analysis shows strong separability for PVH but limited significance for NPVH, our machine learning pipeline, tailored for high-dimensional feature integration, identifies that coupling features are crucial for both tasks. We achieve an AUC of 0.891 for PVH and 0.728 for NPVH, suggesting that while PVH is near-linearly separable, NPVH discrimination benefits from modeling non-linear feature interactions.

Problem

• vocal hyperfunction
• phonotraumatic
• non-phonotraumatic
• biomarkers
• classification

Innovation

• hierarchical feature engineering
• vocal hyperfunction classification
• coupling features
• source-filter interaction
• non-linear feature interactions