Google trains AI on 300M sounds to detect diseases via smartphone audio

In artificial intelligence progress, Google is advancing by employing a novel method that utilizes audio signals to anticipate initial symptoms of sickness.

Google has utilized 300 million audio samples, including coughs, sniffles, and labored breathing, to train its AI foundation model to identify signs of diseases like tuberculosis.

Google has teamed up with Salcit Technologies, an AI startup focused on respiratory healthcare in India, to incorporate this technology into smartphones.

High-risk communities in areas with restricted healthcare access could experience a transformation due to this.

Bioacoustics

Google has previously made efforts to digitize human senses. The company’s investment arm has previously backed startups using AI to identify diseases based on scent.

Exploring bioacoustics, which combines biology and acoustics, demonstrates the growing use of AI to extract important information from the sounds made by humans and animals.

In healthcare, generative AI, the technology behind ChatGPT’s widespread adoption by over 200 million users, is advancing bioacoustics with new capabilities.

Google has developed an AI model called HeAR (Health Acoustic Representations) that utilizes sound signals to anticipate early signs of illness, providing an innovative tool for medical diagnosis.

Easily deployable on smartphones, this technology can track and screen high-risk populations in regions with limited access to costly diagnostic devices such as X-ray machines.

This method’s usefulness is its capability to offer healthcare options in distant areas by utilizing the microphone and AI software integrated into a phone.

Resolving health issues

Tuberculosis is responsible for nearly 4,500 deaths and approximately 30,000 new infections every day, as reported by the World Health Organization.

While tuberculosis is treatable, millions of cases go undiagnosed. In India alone, tuberculosis leads to nearly a quarter-million deaths annually, highlighting the importance of early detection.

Google’s AI was trained using a massive dataset of 300 million audio clips, which included coughs and breathing sounds from all over the world.

These sounds were obtained from publicly available, non-copyrighted materials, such as YouTube videos and recordings of TB screenings in hospitals in Zambia.

The AI tool, integrated into a smartphone, can be taken to the most remote locations to screen for the disease.

By analyzing subtle differences in cough patterns, the AI system can identify early signs of tuberculosis, facilitating early intervention and treatment.

Google’s partnership with Salcit Technologies aims to enhance the accuracy of tuberculosis diagnosis and lung health assessments.

Swaasa

Salcit is merging Google’s AI model with its machine learning technology, Swaasa, an AI system named after the Sanskrit word for breath.

This collaboration is expected to greatly enhance the monitoring of respiratory health and the management of diseases, especially in areas with limited access to healthcare professionals and diagnostic tools.

The use of AI to detect diseases through sound represents a significant technological breakthrough with the potential to revolutionize healthcare delivery.

As AI models such as HeAR become more advanced, they could expand beyond detecting tuberculosis to identifying other respiratory illnesses and cardiovascular conditions through sound analysis.

Developing such tools is crucial in a world where healthcare accessibility remains challenging for millions.

Utilizing smartphones’ current infrastructure, these AI-based solutions can be quickly expanded and used in urban and rural areas, improving the inclusivity and accessibility of healthcare.