Early Detection of Tuberculosis with Machine Learning Cough Audio Analysis: Towards More Accessible Global Triaging Usage

Chandra Suda

2023-10-26

Abstract:Tuberculosis (TB), a bacterial disease mainly affecting the lungs, is one of the leading infectious causes of mortality worldwide. To prevent TB from spreading within the body, which causes life-threatening complications, timely and effective anti-TB treatment is crucial. Cough, an objective biomarker for TB, is a triage tool that monitors treatment response and regresses with successful therapy. Current gold standards for TB diagnosis are slow or inaccessible, especially in rural areas where TB is most prevalent. In addition, current machine learning (ML) diagnosis research, like utilizing chest radiographs, is ineffective and does not monitor treatment progression. To enable effective diagnosis, an ensemble model was developed that analyzes, using a novel ML architecture, coughs' acoustic epidemiologies from smartphones' microphones to detect TB. The architecture includes a 2D-CNN and XGBoost that was trained on 724,964 cough audio samples and demographics from 7 countries. After feature extraction (Mel-spectrograms) and data augmentation (IR-convolution), the model achieved AUROC (area under the receiving operator characteristic) of 88%, surpassing WHO's requirements for screening tests. The results are available within 15 seconds and can easily be accessible via a mobile app. This research helps to improve TB diagnosis through a promising accurate, quick, and accessible triaging tool.

Audio and Speech Processing,Machine Learning,Sound,Quantitative Methods