AI is revolutionizing healthcare, with companies like Google leading the way by developing bioacoustic models that detect diseases from human sounds. Recently, Google partnered with India’s Salcit Technologies to create an AI model, "HeAR" (Health Acoustic Representations), trained on millions of audio clips, including coughs and sneezes. This technology allows smartphones to act as portable diagnostic tools, especially in areas lacking access to advanced medical equipment.
Tuberculosis (TB), which kills nearly 250,000 people yearly in India, is a primary focus. Google’s AI listens to subtle cough differences, helping to screen for TB. Salcit’s Swaasa app further enhances this, allowing users to upload a 10-second cough sample for analysis with 94% accuracy. This low-cost test could potentially replace expensive procedures like spirometry.
However, challenges remain, such as ensuring clean audio samples in rural settings. Despite this, the technology is gaining support from organizations like the StopTB Partnership.
Beyond TB, Google is exploring AI for early breast cancer detection through ultrasound. While these models aren’t yet commercially available, sound-based AI holds promise for making disease detection more accessible and affordable globally.
Source
Photo: Getty