Accelerating cough-based algorithms for pulmonary tuberculosis screening: Results from the CODA TB DREAM Challenge

Background Open-access data challenges can accelerate innovation in artificial-intelligence (AI)-based tools. In the Cough Diagnostic Algorithm for Tuberculosis (CODA TB) DREAM challenge, we developed and independently validated cough sound-based AI algorithms for TB screening. Methods We included data from 2,143 adults with ≥2 weeks of cough from outpatient clinics in India, Madagascar, the Philippines, South Africa, Tanzania, Uganda, and Vietnam. A standard TB evaluation was completed, and at least three solicited coughs were recorded using a smartphone. We invited teams to develop models using training data to classify microbiologically confirmed TB disease using 1) cough sound features only and/or 2) cough sound features with routinely available clinical data. After four months, they submitted the algorithms for independent test set validation. Models were ranked by area under the receiver operating characteristic curve (AUROC) and partial AUROC (pAUROC) to achieve at least 80% sensitivity and 60% specificity. Results Eleven cough models and six cough-plus-clinical models were submitted. AUROCs for cough models ranged from 0.69-0.74, and the highest performing model achieved 55.5% specificity (95% CI 47.7-64.2) at 80% sensitivity. The addition of clinical data improved AUROCs (range 0.78-0.83); five of the six models reached the target pAUROC, and the highest performing model had 73.8% (95% CI 60.8-80.0) specificity at 80% sensitivity. The AUROC varied by country, and was higher among males and HIV-negative individuals. Conclusions In a short period, an open-access data challenge facilitated the development of new cough-based TB algorithms and demonstrated potential as a TB screening tool.