A web tool that incorporates artificial intelligence to help people recognize skin, hair, and nail ailments is Google’s new move into health care – Google AI health tool for skin conditions. The tool was shown off today at Google I/O, and the company says it plans to start a pilot later this year.
People may use the camera on their phone to take three photos of the problem spot, such as a rash on their arm. After that, they’ll be asked a series of details about their skin type as well as other symptoms. The tool then generates a list of potential conditions from a database of 288 conditions that it has been trained to identify. In a blog post, the company stated that it is not designed to diagnose the issue.
Because of the existence of skin conditions, Google opted to use artificial intelligence to combat them, according to Karen DeSalvo, Google Health’s chief health officer. “People are coming to Google to ask questions about skin conditions. We get about 10 billion annual skin condition queries,” she stated.
Experts can certainly assist patients in determining if their symptoms are minor or indicative of a more severe disease, but dermatologists are in short supply all over the world. DeSalvo hopes that by using this method, people will be able to easily obtain reliable information about possible illnesses without having to spend hours doing their own online research.
Millions of images of skin conditions, thousands of images of healthy skin, and 65,000 images from clinical settings were used to train the model. When predicting potential conditions, the model considers factors such as age, skin type, sex, and race. It is claimed that the Google AI health tool for skin conditions correctly identified the right condition in the top three recommendations 84 percent of the time when it was checked on about 1,000 images of skin problems from a wide range of patients. 97 percent of the time, it mentioned the right condition as one of the potential issues.
Google AI health tool builds on Google’s previous work identifying skin conditions using artificial intelligence software. Last spring, the company released the first version of its deep learning framework in Nature Medicine. The system could recognize 26 common skin conditions as accurately as dermatologists and more accurately than primary care physicians, according to the report. Another research published by the company in April found that the device could aid non-dermatologist physicians in making more accurate diagnoses of skin conditions.