Spasmodic Dysphonia Detection Using Machine Learning Classifiers

Abstract

Spasmodic Dysphonia (SD) is a neurological problem that involves the laryngeal muscles to malfunction. It is characterized by inappropriate contraction of the laryngeal muscles during speech. To distinguish healthy and pathological human voices, we used a variety of machine learning classifiers to conduct a side-by-side comparison for the detection of this vocal problem. Three commonly used classifiers, namely, knearest neighbors (KNN), Support Vector Machine (SVM), and Decision Tree (DT) were used in this study. Our study is based on the Saarbruecken Voice Database (SVD), a freely accessible German database that contains various samples, vowels, and sentences from normal and diseased voices. In this article, we solely used the sustained phonation of the vowel /a/ low pitch, and the DT algorithm yielded better classification accuracy of roughly 86.66 %.