Home / Health / AI coughs up COPD diagnosis
AI coughs up COPD diagnosis
19 Feb
Summary
- AI app screens for COPD using cough sounds on smartphones.
- Validation study by AIIMS showed strong agreement in severe cases.
- App is 90% accurate distinguishing normal from abnormal cases.
A new artificial intelligence-powered mobile application, Shwaasa, is poised to enhance primary healthcare diagnostics in India. Developed by a Karnataka-based startup, this app utilizes an algorithm-based platform to screen individuals for Chronic Obstructive Pulmonary Disease (COPD), a significant cause of illness in the country.
AIIMS-Delhi conducted a validation study on 460 participants, comparing the app's performance with spirometry. The findings indicated a moderate overall correlation and strong agreement, particularly in severe COPD cases. The app analyzes cough sounds captured by a smartphone's microphone, providing diagnostic output within eight minutes.
This innovative tool boasts approximately 90% accuracy in differentiating normal from abnormal lung conditions, with specific accuracy rates for COPD and asthma ranging from 82% to 87%. Its deployment is recommended in primary and secondary healthcare facilities where spirometry, the gold standard for lung function testing, is often inaccessible due to cost and equipment limitations.
Unlike spirometry, Shwaasa requires no specialized training or elaborate equipment, making it suitable for resource-limited settings. AIIMS doctors suggest its use in facilities like Ayushman Arogya Mandirs as an effective community-level screening tool. Additionally, AIIMS is exploring its utility for tuberculosis screening.
