Google’s latest research in healthcare is an online tool that uses artificial intelligence to help people identify skin, hair, or nail conditions. The company previewed the tool in I / O today and says it hopes to launch the pilot later this year.
People can use the phone’s camera to take three pictures of the problem area – for example, a rash on their arm. They then answer several questions about skin type and other symptoms. The tool then provides a list of possible conditions out of 288 that it has been trained to identify. It is not intended to diagnose the problem, the company said blog post.
Google decided to infect skin diseases with artificial intelligence because of their prevalence, says Karen DeSalvo, health director at Google Health. “People are coming to Google to inquire about skin conditions. We get about 10 billion annual demand for skin disease,” he said in an interview Limit. Of course, experts can help people figure out if it’s something simple or indicates a more serious illness, but there is a shortage of dermatologists all over the world. DeSalvo hopes this tool can help people get accurate information about possible conditions quickly without having to spend hours doing their 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 of clinical conditions. The model takes into account factors such as age, skin type, gender, and race when proposing possible conditions. When it was tested with about 1,000 images of skin problems from a variety of patients, Google says it identified the right condition in the first three proposals 84 percent of the time. It contained proper condition as one of the potential problems 97 percent of the time.
The new system builds on Google’s previous work using artificial intelligence tools to identify skin diseases. The company released the first iteration of its deep learning system in Natural medicine last spring. That paper showed that the system was able to identify 26 common skin conditions as accurately as dermatologists and more accurately than primary care physicians. In April, the company released another study shows that the system could help non-dermatologists diagnose skin diseases more accurately.
Google is also testing with Stanford University’s research team how well the tool works in healthcare.
The company obtained a Class I medical device label for the tool in the European Union, which meant it as a low-risk medical device. It has not been evaluated by the US Food and Drug Administration.
Developing … we’ll add more to this post, but you can follow our post Google I / O 2021 Blog to get the news even faster.