Mechanism-Specific Speech Encoding Failures in Auditory Neuropathy: A Computational Phenotyping Framework

Auditory Neuropathy Spectrum Disorders (ANSD) are best characterized by distorted patterns of auditory nerve activity despite preserved spectral analysis of sound in the cochlea. They should clinically translate into impaired speech recognition despite normal auditory sensitivity. While distinct pathophysiological mechanisms affecting auditory-nerve activity have been identified in animal models, current clinical speech tests cannot distinguish among them. Our working hypothesis is that current speech audiometry yields aggregate recognition scores that average across phoneme categories, obscuring mechanism-specific patterns instead of pinpointing distinctive signatures. Using computational modeling of auditory nerve responses, we tested four mechanism types, demonstrating that mechanism-specific encoding disruptions cascade into speech recognition failures. Brief consonants showed severe disruption while sustained vowels were preserved, with category-specific patterns differing across mechanisms. Models trained on ANSD-degraded signals developed compensation strategies that generalized to healthy signals, while the reverse failed completely: strategies exploiting slow temporal structure (sustained formants) generalize, while those requiring millisecond-scale timing do not. Noise training that benefited healthy models harmed ANSD models, explaining real-world listening difficulties. Phoneme-specific confusion patterns enable mechanism identification from behavioral testing alone, providing the missing diagnostic infrastructure for targeted intervention.