Predicting Tuberculosis from Real-World Cough Audio Recordings and Metadata


Results suggest mobile phone-based applications that integrate clinical symptoms and cough sound analysis could help community health workers and, most importantly, health service programs to improve TB case-finding efforts while reducing costs, which could substantially improve public health.

This abstract focuses on using cough audio recordings and related metadata to predict tuberculosis (TB). This study involves collecting cough sounds from individuals and analyzing various characteristics of these sounds alongside metadata. The aim is to determine patterns or indicators that are specific to TB, thus enabling early and non-invasive diagnosis. This approach would represent a significant advancement in TB diagnosis, offering a more accessible and potentially cost-effective method compared to traditional diagnostic techniques. The research's outcome could have substantial implications for TB detection, particularly in regions where access to medical facilities is limited.