About
Every year about 16000 people look for medical attention for breath issues and copd, with this app they can use their mobile phone to upload a sample of their breath , cough and some words, and they receive a signal if there is an issue.
This project delivers a zero-cost, instant respiratory and cardiac early warning check accessible to anyone with a smartphone. Using only the device microphone, the web app guides users through a short, clinically inspired assessment—breathing, coughing, and speech tasks—and applies lightweight AI to detect patterns associated with conditions such as Chronic Obstructive Pulmonary Disease, asthma, and early respiratory infections. In under two minutes, users receive a simple risk indication (green/amber/red), empowering them to seek care earlier, reduce uncertainty, and avoid deterioration before symptoms become critical. The solution is fully browser-based, requires no download, and works on any modern phone—making it inherently scalable across the ~3 million population served by NHS Wales.
The impact is immediate and system-wide: by shifting detection earlier in the pathway, this tool has the potential to reduce avoidable GP visits, prevent emergency admissions, and support proactive self-management—particularly for high-risk and underserved populations. It complements existing NHS services by acting as a front-door triage layer, filtering low-risk cases while flagging those needing urgent attention. Built with privacy in mind (audio processed transiently, no long-term storage), and deployable at near-zero marginal cost, the platform demonstrates how AI can be responsibly embedded into everyday care. With further validation, it could integrate into NHS digital pathways, community outreach, and remote monitoring programmes—turning any smartphone into a scalable, equitable screening tool for respiratory health.
Presentation Video
Code and Licence
Source code: https://github.com/gbucci1972/BreathCheckAI