A Review and Classification of Amyotrophic Lateral Sclerosis with Speech as a Biomarker
Shaik Mulla Shabber, Mohan Bansal, Kodali Radha
Abstract
Amyotrophic Lateral Sclerosis (ALS) is a motor system neurodegenerative disease that affects speech impairment, spinal, respiratory and swallowing difficulties in patients. It has gradually increased in elderly people in recent years and is not easy to diagnose. The ALS bulbar form system is based on detecting dysarthria speech classification in discriminating healthy subjects from ALS patients. To construct the classification model by using various machine learning techniques, the studies used datasets related to speech impairment recordings of ALS patients and healthy subjects. Early diagnosis of ALS can somewhat improve the patient’s quality of life to an extent. Sustained vowel phonation is very useful for classifying ALS and healthy control (HC). In this study, jitter and shimmer were used to extract features. A support vector machine classifier was applied, and it classified ALS/HC with an accuracy of 98.5%. This study presents a detailed review of various machine learning techniques applied to the speech signal for the diagnosis of ALS and their impact on future research in this direction.