Publication:

Diagnostic accuracy of the TimBre acoustic device for detecting pulmonary tuberculosis: a systematic review and meta-analysis

Tuberculosis (TB) remains a major cause of infectious disease mortality worldwide, with India accounting for 26% of global cases. Conventional screening methods often miss asymptomatic or minimally symptomatic cases, highlighting the need for rapid, non-invasive detection tools. The TimBre acoustic system uses machine learning algorithms to analyze cough sounds for pulmonary TB (PTB) screening. A systematic review and meta-analysis were conducted following PRISMA 2020 guidelines to evaluate TimBre’s diagnostic accuracy. Studies comparing TimBre with reference standards such as Xpert MTB/RIF, Xpert Ultra, TrueNat, or culture were identified from PubMed, Google Scholar, company sources, and grey literature. Pooled sensitivity and specificity were calculated using a random-effects model, and study quality was assessed with QUADAS-2. Three studies met inclusion criteria, encompassing populations from India, Vietnam, the Philippines, Uganda, and South Africa, with over 700,000 cough recordings analyzed. The pooled sensitivity was 71% (95% CI: 68-75%) and pooled specificity 76% (95% CI: 60-87%), with notable heterogeneity across studies. TimBre demonstrates moderate diagnostic performance as a rapid, low-cost, and non-invasive PTB screening tool. Its scalability and ease of deployment make it a promising triage solution for early TB detection, particularly in resource-limited settings. Larger multi-center validations and cost-effectiveness studies are warranted to guide programmatic integration.