Tuberculosis Screening from Cough Audio: Baseline Models, Clinical Variables, and Uncertainty Quantification

TL;DR

This work establishes a reproducible baseline pipeline for TB screening from cough audio augmented with clinical data, addressing methodological variance across prior studies. It combines carefully engineered acoustic features with simple baselines (Logistic Regression) and a strong tabular model (CatBoost), under a cougher-disjoint nested cross-validation framework and with model-agnostic uncertainty quantification via conformal prediction. The results show that multimodal fusion substantially improves discrimination (ROC AUC ≈ 0.80–0.81) and yields reliable probability calibration, while conformal prediction provides actionable uncertainty outputs and selective decision options. By releasing the full experimental protocol and code, the study offers a robust reference point to enable fair benchmarking and future improvements focused on modeling rather than data handling, ultimately supporting uncertainty-aware TB screening from cough in real-world settings.

Abstract

In this paper, we propose a standardized framework for automatic tuberculosis (TB) detection from cough audio and routinely collected clinical data using machine learning. While TB screening from audio has attracted growing interest, progress is difficult to measure because existing studies vary substantially in datasets, cohort definitions, feature representations, model families, validation protocols, and reported metrics. Consequently, reported gains are often not directly comparable, and it remains unclear whether improvements stem from modeling advances or from differences in data and evaluation. We address this gap by establishing a strong, well-documented baseline for TB prediction using cough recordings and accompanying clinical metadata from a recently compiled dataset from several countries. Our pipeline is reproducible end-to-end, covering feature extraction, multimodal fusion, cougher-independent evaluation, and uncertainty quantification, and it reports a consistent suite of clinically relevant metrics to enable fair comparison. We further quantify performance for cough audio-only and fused (audio + clinical metadata) models, and release the full experimental protocol to facilitate benchmarking. This baseline is intended to serve as a common reference point and to reduce methodological variance that currently holds back progress in the field.

Figures (2)

• Figure 1: Cougher-disjoint nested CV pipeline for model selection, calibration, and conformal prediction based uncertainty quantification.
• Figure 2: MFCC and Chroma features for two cough waveforms, TB+ and TB-.