Detecting COPD Through Speech Analysis: A Dataset and Machine Learning Approach
Term
4. Term
Publication year
2025
Submitted on
2025-06-05
Pages
10
Abstract
This study explores the application of voice recordings and machine learning to support early detection of Chronic Obstructive Pulmonary Disease (COPD). Audio data were collected from 96 participants through three vocal tasks, and features were extracted using openSMILE and SpeechBrain. Four models were tested across multiple data configurations. Results show that SVM and Random Forest models performed consistently well, especially with openSMILE features. While limitations include reliance on self-reported diagnoses and inconsistent task execution, the findings suggest that voice-based analysis has potential as a non-invasive, scalable screening tool for COPD.
Documents