Artificial Intelligence as Medical Device (AIaMD) Solution for Primary Care: Transforming Smartphones into Diagnostic Stethoscopes for Telemedicine in Respiratory Health

Our innovation is a groundbreaking AI as Medical Device (AIaMD) that transforms any smartphone into a sophisticated stethoscope, designed to enhance primary care telemedicine. It uniquely applies artificial intelligence to analyse respiratory sounds, facilitating the telemedicine by real time analysis, interoperability, remote monitoring and follow-up of conditions such as asthma (8 million in the UK), COPD (1.2 million), flu, COVID-19, and post-COVID syndrome in a General Practise setting [1-5]. Our AI algorithm expertly filters ambient noise, distinguishing between normal and abnormal respiratory sounds, thereby addressing a significant unmet health need - the lack of physical examination tools in tele-health consultations [6-8]. Currently we completed TRL 3, we have developed proof of concept of both software and hardware components, however, subsequent trials revealed that our software can effectively filter and analyse body sounds using just a smartphone, eliminating the need for additional hardware. Initial trials demonstrated that our software is capable of capturing medical-grade body sounds through common mobile devices. This finding has pivoted our focus from hardware development to enhancing the software's AI capabilities and expanding its accessibility. Patient and public involvement has been central to our development process. We conducted survey with both patients and healthcare professionals to assess the need for respiratory sound analysis in telemedicine. Feedback indicated the critical nature of this feature for managing the aforementioned conditions. Respondents also highlighted the importance of inclusivity, prompting us to expand our user interface to support multiple languages, ensuring accessibility for diverse and underserved communities. With the support of i4i FAST funding, we aim to advance our technology to TRL 4, validating our AI algorithm in a controlled laboratory setting, which is a critical step towards its clinical application in follow up telemedicine consultations in Primary Care and General Practise settings.