Comparison of Speaker Role Recognition and Speaker Enrollment Protocol for conversational Clinical Interviews

TL;DR

This study compares speaker role recognition and speaker enrollment methods using neural networks on clinical conversations, finding that role recognition performs best and emphasizing the importance of domain-specific retraining.

Contribution

It provides a comparative evaluation of speaker identification methods in clinical settings, highlighting the effectiveness of speaker role recognition models.

Findings

Speaker role recognition outperforms speaker enrollment.

Retraining models with in-domain data improves performance.

Results are consistent across different patient demographics.

Abstract

Conversations between a clinician and a patient, in natural conditions, are valuable sources of information for medical follow-up. The automatic analysis of these dialogues could help extract new language markers and speed-up the clinicians' reports. Yet, it is not clear which speech processing pipeline is the most performing to detect and identify the speaker turns, especially for individuals with speech and language disorders. Here, we proposed a split of the data that allows conducting a comparative evaluation of speaker role recognition and speaker enrollment methods to solve this task. We trained end-to-end neural network architectures to adapt to each task and evaluate each approach under the same metric. Experimental results are reported on naturalistic clinical conversations between Neuropsychologist and Interviewees, at different stages of Huntington's disease. We found that our Speaker Role Recognition model gave the best performances. In addition, our study underlined the importance of retraining models with in-domain data. Finally, we observed that results do not depend on the demographics of the Interviewee, highlighting the clinical relevance of our methods.