(SimPhon Speech Test): A Data-Driven Method for In Silico Design and Validation of a Phonetically Balanced Speech Test

2025-06-13

AI Summary

Conventional hearing tests inadequately characterize suprathreshold speech understanding deficits commonly observed in presbycusis. To address this, we propose the first computational framework for speech testing that leverages automatic speech recognition (ASR) to model hearing-impaired perception—enabling end-to-end automation of confusion pattern mining, corpus-driven minimal-pair generation, and expert-in-the-loop optimization. Innovatively, we employ ASR systems as perceptual surrogates for hearing impairment, yielding SimPhon-25: a phonemically balanced, highly discriminative test set comprising 25 minimal word pairs. Validation demonstrates no significant correlation (*p* > 0.05) between SimPhon-25 scores and the traditional Speech Intelligibility Index (SII), confirming its specificity in capturing non-audibility-based deficits. The framework substantially improves development efficiency and is now ready for Phase I clinical trials.

Technology Category

• Natural Language Processing: Speech
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Traditional audiometry often provides an incomplete characterization of the functional impact of hearing loss on speech understanding, particularly for supra-threshold deficits common in presbycusis. This motivates the development of more diagnostically specific speech perception tests. We introduce the Simulated Phoneme Speech Test (SimPhon Speech Test) methodology, a novel, multi-stage computational pipeline for the in silico design and validation of a phonetically balanced minimal-pair speech test. This methodology leverages a modern Automatic Speech Recognition (ASR) system as a proxy for a human listener to simulate the perceptual effects of sensorineural hearing loss. By processing speech stimuli under controlled acoustic degradation, we first identify the most common phoneme confusion patterns. These patterns then guide the data-driven curation of a large set of candidate word pairs derived from a comprehensive linguistic corpus. Subsequent phases involving simulated diagnostic testing, expert human curation, and a final, targeted sensitivity analysis systematically reduce the candidates to a final, optimized set of 25 pairs (the SimPhon Speech Test-25). A key finding is that the diagnostic performance of the SimPhon Speech Test-25 test items shows no significant correlation with predictions from the standard Speech Intelligibility Index (SII), suggesting the SimPhon Speech Test captures perceptual deficits beyond simple audibility. This computationally optimized test set offers a significant increase in efficiency for audiological test development, ready for initial human trials.

Problem

• Develops a phonetically balanced speech test for hearing loss diagnosis
• Uses ASR to simulate hearing loss effects on phoneme perception
• Creates optimized word pairs uncorrelated with standard audibility metrics

Innovation

• Uses ASR to simulate hearing loss effects
• Data-driven phoneme confusion pattern analysis
• Optimized 25-pair test via computational pipeline