Google has announced a strategic partnership with Salcit Technologies, an Indian respiratory healthcare firm, to enhance the early detection of tuberculosis (TB) using artificial intelligence (AI). This collaboration centres around integrating Google's HeAR (Health Acoustic Representations) AI model with Salcit's innovative Swaasa platform, which analyzes cough sounds to assess lung health. The initiative aims to improve the accuracy and efficiency of TB diagnostics significantly.
The partnership, unveiled by Google on August 20, underscores the transformative potential of AI in respiratory healthcare. Google's HeAR model, a bioacoustic foundational AI system, has been trained on a vast dataset of 300 million audio samples, including around 100 million cough sounds. This extensive training enables the model to detect subtle variations in cough acoustics, which are critical for early and precise TB diagnosis.
Shravya Shetty, Director and Engineering Lead at Google Health, highlighted the potential impact of HeAR during a media roundtable. She emphasized that the model's ability to identify minute differences in cough sounds could significantly improve the early detection of TB. This capability allows for earlier patient triaging and prioritization of follow-up testing, which is crucial for the timely confirmation and treatment of tuberculosis cases.