Machine learning for amplitude-modulation depth discrimination tests: Evaluation of validity and accuracy

Abstract

This proposal will evaluate a new automated method for measuring the amplitude-modulation depth discrimination threshold. Using a machine learning method to measure the threshold will be assessed and proven to be an accurate, reliable, and time efficient tool. Machine learning is accurate, as it measures the threshold on a continuous frequency scale and requires much less time than the conventional method. The findings of this study will help improve the HAs experience for those with hearing loss. This faster and accurate way to measure the amplitude-modulation depth discrimination threshold will enable more frequent testing in audiology clinics. Because this test will provide more information about an individual’s hearing loss, those requiring HAs will experience a more suitable and appropriate HAs fitting. Moreover, this method will improve the experience and knowledge of audiologists.