AI Cough Analysis Spots Hidden Respiratory Illnesses

In a groundbreaking pilot in Andhra Pradesh's East Godavari district, an AI-powered platform named Swaasa has begun screening for respiratory illnesses using smartphone technology. Healthcare workers are now asking villagers to cough into a mobile app that analyzes the sounds to identify potential cases of Tuberculosis, COPD, and asthma. This initiative, part of the MedTech Innovation Challenge 2025, successfully screened close to 8,000 people within six weeks. Significantly, about 36 percent of identified TB cases were asymptomatic, highlighting the platform's advantage over traditional screening methods. The AI system demonstrated high concordance with spirometry tests, offering a non-invasive, radiation-free solution.

The deployment integrated into public health workflows enabled door-to-door screening, instant risk stratification, and targeted referrals. Real-time monitoring dashboards helped track progress and identify operational challenges. The project also revealed a notable prevalence of chronic respiratory illnesses, with COPD risk percentages ranging between 6.5 and 9.5, and asthma risk between 1.6 and 1.9 percent. This approach aims to improve TB diagnostic yield by approximately 15 percent. Further integration with national health systems like Ayushman Bharat Digital Mission is planned for wider adoption.