# What Does a Pathological Speech Assessment Model Know about Acoustic Features? A Case Study on Oral and Oropharyngeal Cancer Patients

arXiv: https://arxiv.org/abs/2606.24949 · HTML: https://mlsanity.com/paper/2606.24949/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 34 references that are in the map.
- First posted on arXiv 23 Jun 2026.

## Builds on (most cited first, 8 of 8)

- [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://mlsanity.com/paper/2006.11477/) (2020)
- [Insights on representational similarity in neural networks with canonical correlation](https://mlsanity.com/paper/1806.05759/) (2018)
- [Layer-wise Analysis of a Self-supervised Speech Representation Model](https://mlsanity.com/paper/2107.04734/) (2021)
- [Comparative layer-wise analysis of self-supervised speech models](https://mlsanity.com/paper/2211.03929/) (2022)
- [Wav2vec-based Detection and Severity Level Classification of Dysarthria from Speech](https://mlsanity.com/paper/2309.14107/) (2023)
- [Exploring Pathological Speech Quality Assessment with ASR-Powered Wav2Vec2 in Data-Scarce Context](https://mlsanity.com/paper/2403.20184/) (2024)
- [Exploring ASR-Based Wav2Vec2 for Automated Speech Disorder Assessment: Insights and Analysis](https://mlsanity.com/paper/2410.08250/) (2024)
- [Investigation of Self-supervised Pre-trained Models for Classification of Voice Quality from Speech and Neck Surface Accelerometer Signals](https://mlsanity.com/paper/2308.03226/) (2023)

## Abstract (arXiv)

This work investigates the interpretability of a Wav2Vec 2.0based speech intelligibility assessment model for oral and oropharyngeal cancer patients through canonical correlation analysis. By measuring the correlation between the model embeddings and eGeMAPS low-level descriptors (LLDs) as an interpretable reference, we analyze how acoustic information is encoded across the model layers. The analysis is conducted at two levels: individual LLDs layer-wise, and group-level: prosodic, spectral, and voice quality. Results show that the learned representations are most strongly correlated with spectral and prosodic features, with the first MFCC coefficient yielding the highest correlations across all layers. At the group level, spectral and prosodic groups achieve correlations of 0.77 and 0.71 respectively, while voice quality reaches 0.65. Beyond model interpretability, this work also offers practical guidance on acoustic feature selection for pathological speech assessment.