Evaluation of Disease Severity by Acoustic Analysis of the Eustachian Tube Insufflation Sound

Diseases associated with the Eustachian tube include Eustachian tube stenosis, Eustachian tube opening, and otitis media with effusion. The Eustachian tube insufflation method is used both for examining and treating Eustachian tubes and is clinically performed on patients with Eustachian tube diseases. However, the Eustachian tube insufflation sounds that are produced by this method are only auscultated by the doctor. Therefore, there is a disadvantage that the sound is not saved as data, and diagnosis is based on the auditory judgment of doctors, leading to weak objectivity. The purpose of this study is to improve the objectivity of diagnosis by developing a system that quantitatively evaluates the severity of the Eustachian tube insufflation sounds. We collected Eustachian tube insufflation sounds from patients with the Eustachian tube disease. The acoustic characteristics of the collected data were analyzed in order to quantify the difference in the Eustachian tube insufflation sound, depending on the degree of the Eustachian tube disease. In addition, spectrum analysis was performed using the maximum entropy method (MEM), and the abnormal sound was extracted from the cross-correlation function with the template of the abnormal sound. The change of the frequency spectrum with time was expressed. As a result, the time taken to open the Eustachian tube was recorded and an index indicating the severity of normal and stenotic sounds was proposed.