Google has decided to expand the capabilities of modern artificial intelligence systems – the company is now working on technology that will help detect early signs of diseases based on the sounds made by a person.
Image Source: Pixabay.com
To achieve this goal, Google trained a basic AI model on 300 million fragments of audio recordings that include coughs, signs of a runny nose and difficulty breathing – this initial data will help identify a person who is sick, for example, with tuberculosis, reports Bloomberg. The model is called HeAR (Health Acoustic Representations).
The search giant has also joined forces with Indian company Salcit Technologies. A startup specializing in AI technologies in the field of respiratory diseases will help deploy this technology on smartphones – it will help citizens belonging to high-risk groups in regions with insufficient access to healthcare services.
This isn’t Google’s first attempt to digitally express the human senses. The company’s core division previously supported a startup that develops AI capable of identifying diseases by smell.