Google develops bioacoustic healthcare AI model 'HeAR' that can detect tuberculosis by analyzing a patient's cough
Google's research organization, Google Research, has announced Health Acoustic Representations (HeAR), an AI model that can detect early symptoms of diseases such as tuberculosis by training it on 100 million cough sounds.
[2403.02522] HeAR -- Health Acoustic Representations
Sounds our bodies make — like coughs — contain valuable health information. Learn how AI models built using HeAR, a bioacoustic foundation model trained on 100 million cough sounds, can help flag early signs of diseases like tuberculosis ↓ https://t.co/Mrqgspa0vm
— Google (@Google) August 22, 2024
Can we hear disease before we see it? - YouTube
Researchers built an AI model to detect diseases based on coughs
https://blog.google/technology/health/ai-model-cough-disease-detection/
Google and Others Are Developing AI That Can Hear Signs of Sickness - Bloomberg
https://www.bloomberg.com/news/newsletters/2024-08-29/google-and-others-are-developing-ai-that-can-hear-signs-of-sickness
Developed in partnership with Indian AI startup Salcit Technologies , HeAR was trained on over 300 million sniffles, coughs, sneezes and breathing sounds from Zambian hospitals and YouTube videos, with around 100 million cough sounds included in the training data.
HeAR uses information from countless data to identify health-related sound patterns. According to Google, HeAR outperforms similar models in a variety of tasks and has been shown to have superior performance in capturing patterns in health-related acoustic data. Google says, 'HeAR is an important component in the world of healthcare research, where data is often scarce.'
In addition, Google reported that it is working on research and strengthening early detection of tuberculosis based on cough sounds by combining HeAR with ' Swaasa ' developed by Salcit Technologies. Swaasa is a model that uses AI to analyze the sounds of a patient's cough and evaluate the health of the lungs, and can diagnose the disease with approximately 94% accuracy using 10 seconds of recorded data.
Once considered an incurable disease, tuberculosis has become treatable thanks to advances in medical technology. However, millions of patients around the world are left without access to healthcare services every year due to poverty and other issues. Organizations such as the WHO are
'It's heartbreaking every time we see a tuberculosis diagnosis delayed for any reason. But acoustic biomarkers like HeAR have the potential to put an end to this sad story. We're deeply grateful for the role HeAR can play in our work to eradicate tuberculosis,' said Sujai Kakarmas, product manager at Google Research. 'Solutions like HeAR enable AI-based acoustic analysis, which can break new ground in tuberculosis screening and detection. This could make it an easily accessible tool for people in poor countries where tuberculosis is prevalent, without imposing a financial burden on them,' said Chin Zhizhen, digital health specialist at StopTB Partnership, a UN-sponsored organization working to eradicate tuberculosis.
'HeAR is a major step forward in acoustic health research. We hope to advance the development of future diagnostic tools and monitoring solutions for tuberculosis, chest, pulmonary and other disease areas, and contribute to improving the health of communities around the world through our research,' Google Research reported.
Related Posts:
