AI-Based A LS Detection Through Speech Analysis

S. Dhanalakshmi1, Ruby Kumari R2,Thanushree V S3,Varsha Dattatreya Bhat4,Vigasini S5

Professor, Dept of CSE, K S Institute of Technology , Karnataka, India1

UG Scholars, Dept of CSE, K S Institute of Technology , Karnataka, India 2,3,4

Abstract: Amyotrophic Lateral Sclerosis(ALS) is a disease that affects the nerves responsible for muscle movement and speech[1][5] . In the early  stages of ALS ,people may experience   changes in their speech such as unclear pronunciation , unusual pauses  and changes in voice pitch[1][3] . Detecting ALS early is difficult because current medical tests are expensive and often identify the disease only after symptoms become severe.  This project aims to develop a deep learning-based system that can help detect ALS early by analysing speech patterns[2][4] . The system collects speech recordings from users over a period of two weeks to observe changes in their voice. The recorded audio is cleaned and processed to improve quality and important speech features such as MFCC, pitch variation , jitter and shimmer are extracted for analysis[1][3] . Deep learning models like CNN and LSTM are then used to predict whether the speech shows signs of ALS or not[2][4] .The proposed system provides a simple, low cost and non-invasive method for early ALS screening and monitoring [5][7] .It can also support remote healthcare by helping users track speech changes over time [5][7] .  Keywords:  Amyotrophic Lateral Sclerosis(ALS), Deep Learning, CNN,  Artificial Intelligence,  Speech Pattern Analysis, Healthcare Monitoring