![[2601.07969] Tuberculosis Screening from Cough Audio: Baseline Models, Clinical Variables, and Uncertainty Quantification](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2601.07969] Tuberculosis Screening from Cough Audio: Baseline Models, Clinical Variables, and Uncertainty Quantification
Summary
This paper presents a standardized framework for tuberculosis detection from cough audio, addressing inconsistencies in previous studies and establishing a reproducible baseline for evaluation.
Why It Matters
The study is significant as it tackles the challenge of varying methodologies in TB screening research, which hinders progress. By providing a clear baseline and protocol, it aims to enhance comparability and reliability in future studies, potentially improving TB detection methods globally.
Key Takeaways
- Introduces a standardized framework for TB detection from cough audio.
- Addresses methodological inconsistencies in existing TB screening studies.
- Provides a reproducible baseline for evaluation and benchmarking.
- Quantifies performance for both audio-only and fused models.
- Releases a full experimental protocol to facilitate future research.
Electrical Engineering and Systems Science > Audio and Speech Processing arXiv:2601.07969 (eess) [Submitted on 12 Jan 2026 (v1), last revised 13 Feb 2026 (this version, v2)] Title:Tuberculosis Screening from Cough Audio: Baseline Models, Clinical Variables, and Uncertainty Quantification Authors:George P. Kafentzis, Efstratios Selisios View a PDF of the paper titled Tuberculosis Screening from Cough Audio: Baseline Models, Clinical Variables, and Uncertainty Quantification, by George P. Kafentzis and 1 other authors View PDF 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 quantif...
