Objective Assessment System for Hearing Prediction Based on Stimulus-Frequency Otoacoustic Emissions

Area of Science:

• Audiology
• Otoacoustic Emissions
• Hearing Science

Background:

• Stimulus-frequency otoacoustic emissions (SFOAEs) offer a noninvasive method for evaluating cochlear function.
• Limited research exists on the predictive value of SFOAEs for hearing capabilities.

Purpose of the Study:

• To develop and validate an objective assessment system based on SFOAEs for predicting hearing thresholds and identifying hearing loss.
• To evaluate the performance of routine and fast testing modules for threshold prediction and a screening module for hearing status identification.

Main Methods:

• Trained predictive models using SFOAE data from 245 normal-hearing ears and 839 ears with sensorineural hearing loss.
• Developed three modules: routine test, fast test (for threshold prediction), and hearing screening.
• Validated model performance using cross-validation and independent unknown datasets.

Main Results:

• Routine and fast test modules demonstrated similar performance in predicting hearing thresholds from 0.5 to 8 kHz (mean absolute errors: 7.06-11.61 dB and 7.40-12.60 dB, respectively).
• The fast module offered reduced testing time compared to the routine module.
• The hearing screening module achieved high accuracy (88.4-95.9%) and a low false-negative rate (2.9-7.0%) across 0.5-8 kHz, with a large area under the ROC curve (0.912-0.985).

Conclusions:

• The developed SFOAE-based system shows potential for objective hearing assessment.
• The routine and fast modules are viable for hearing threshold prediction, though improvement is needed for ears with hearing loss.
• The hearing screening module demonstrates significant promise as a clinical tool for identifying hearing loss.