Google’s latest foray into healthcare is a web-based tool that uses artificial intelligence to help people determine the condition of their skin, hair, or nails. The company announced the tool at I / O today and says it hopes to launch a pilot version later this year.
People can use their phone’s camera to take three pictures of a problem area, such as a rash on their arm. They will then answer a series of questions about their skin type and other symptoms. The tool then lists the possible conditions from a set of 288 that it is trained to recognize. The company stated on its blog that this is not intended to diagnose the problem.
Google has decided to fight skin diseases with artificial intelligence because of their prevalence, said Karen DeSalvo, chief health officer at Google Health. “People come to Google to ask questions about skin conditions. We receive about 10 billion inquiries about the condition of the skin a year, ”she said in an interview. Facets… Of course, experts can help people determine if this is something simple or a sign of a more serious medical condition, but there are not enough dermatologists around the world. DeSalvo hopes this tool will help you quickly get accurate information about potential diseases without spending hours doing your own online research.
The team trained the model on millions of images of skin problems, thousands of images of healthy skin, and 65,000 images from clinical settings. The model considers factors such as age, skin type, gender, and race to suggest possible conditions. When it was tested on roughly 1,000 images of skin problems in a wide variety of patients, Google says that 84 percent of the time it identified the correct condition in three main sentences. He included the correct condition as one of the possible problems 97% of the time.
The new system builds on Google’s past work using artificial intelligence tools to detect skin conditions. The company has published the first version of its deep learning system in Nature Medicine The last spring. This document showed that the system can identify 26 common skin conditions as accurately as dermatologists and more accurately than primary care physicians. In April, the company published another study showing the system could help non-dermatologists diagnose skin conditions more accurately.
Google is also working with a Stanford University research group to test how well the tool works in healthcare settings.
The company has received the Class I Medical Device Label for this instrument in the European Union, designating it as a Low Risk Medical Device. It has not been evaluated by the US Food and Drug Administration.