Salcit Technologies explores TB detection with Google’s AI bioacoustic model

The Economic Times
Aug. 21, 2024, 10:49 p.m.

Aug 21, 2024 -- Search giant Google said on Tuesday that respiratory healthcare startup Salcit Technologies is exploring using its Health Acoustic Representations (HeAR) bioacoustic foundation model to enhance early detection of tuberculosis based on cough sounds.

Trained on about 300 million pieces of audio data, and in particular roughly 100 million cough sounds, HeAR is expected to help the Indian startup extend screening for TB more widely across the country.

“Compared to blood tests and imaging, sound is by far the most accessible piece of information that we can get about a person," Sujay Kakarmath, a product manager at Google Research, said in a statement. "HeAR can pick up chest x-ray findings, tuberculosis and even detect Covid from cough sounds."

"With HeAR, we hope that researchers will be able to discover new acoustic biomarkers a lot faster," Kakarmath said.

Salcit Technologies intends to use HeAR to build on research for its product Swaasa, which has a history of using machine learning to help detect diseases early, bridging the gap with accessibility, affordability and scalability by offering location-independent, equipment-free respiratory health assessment.

Publicly introduced in March 2024, HeAR is designed to help researchers build models that can listen to human sounds and flag early signs of disease.

“We found that, on average, HeAR ranks higher than other models on a wide range of tasks and for generalising across microphones, demonstrating its superior ability to capture meaningful patterns in health-related acoustic data," said Shravya Shetty, director and engineering lead, Health AI, at Google Research.

Shetty said models trained using HeAR also achieved high performance with less training data, a crucial factor in the often data-scarce world of healthcare research.

"Our goal is to enable further research into models for specific conditions and populations, even if data is sparse or if cost or compute barriers exist," she said.

Google invited researchers interested in exploring HeAR's potential to request access to the HeAR API.

Through its research, Google said it hopes to keep helping advance the development of future diagnostic tools and monitoring solutions that assist improved health outcomes for communities around the globe.


Source: The Economic Times