AI-based hearing loss diagnosis method using voice data

Abstract

This invention provides a method for diagnosing the presence or absence of hearing loss early, simply, and accurately by using only the subject's voice data and an artificial intelligence (AI) algorithm. The method utilizes voice data from reading negative prose (long, scenario-style texts approximately one minute in length) and positive prose. Specifically, it employs voice data from reading negative prose with full emotion (such as regret or intense feelings) and positive prose without emotion. This allows for accurate and effective diagnosis of hearing loss regardless of the subject's age or gender. Furthermore, for voice data that is difficult to obtain sufficiently due to collection processes or security issues, the invention uses expanded data in conjunction with the original data to reduce overfitting due to the amount of data, while effectively utilizing the acquired voice data to accurately diagnose hearing loss. The invention also identifies effective voice features and diagnostic algorithms for early diagnosis of hearing loss from the acquired voice data, enabling a more effective, early, simple, and accurate diagnosis of hearing loss using AI-based voice data.

Images