Google announced a new tool that helps its users to detect skin diseases, through the company´s artificial intelligence chips, and it is scheduled to be launched later this year.
Peggy Boy, director of the new tool at Google Health, says that there are many global search engine users looking for information about skin, hair and nail problems, with up to 10 billion queries every year, but they do not have access to accurate information most of the time.
Bowie added, "Almost two billion people around the world suffer from skin diseases, and most often the first step for patients is to resort to (Google), but words cannot properly describe the situation, which leads to inaccurate information."
How to use
During the "Google" conference this week, the company reviewed the new tool to help detect skin diseases, and made it clear that the user uploads three well-lit and clear images of the skin, hair or nails from different angles, after which the tool raises a series of questions about the user´s skin type. How long the problem has occurred and other symptoms that help the tool narrow the possibilities.
The next step is an analysis of the tool´s artificial intelligence model for that information, based on various symptoms of skin diseases, which have been reviewed by specialized doctors, and according to Peggyboy, the tool is based on 288 conditions to provide a list of possible match conditions between the user´s disease condition and the expectation of the tool. to her.
Next, performance provides dermatologist-reviewed information and answers to common questions, along with identical and similar images. Users can either save or delete their results or contribute them to Google´s research and development.
Also, the new product manager at Google confirms that the information provided by users is stored and encrypted securely, and that the company does not use that data to target ads.
Bowie notes that the tool will help users to initially recognize the skin disease, but it cannot be considered a substitute for doctors.
business trip
It took three years of machine learning research and development to produce that new tool, according to the project´s technical leader, Yuan Liu.
So far, Google has published many of the research papers that have been reviewed, and one of its role is to validate the artificial intelligence model, and there is more research in progress.
In a study published in the scientific journal Nature, last year, it was found that the AI ​​system is as good as a dermatologist at identifying 26 skin conditions, and is more accurate than primary care doctors and nurses.
In the same context, a recent research paper published in the journal "Gamma Network Open" showed that the new tool may help doctors diagnose skin diseases more accurately during an initial evaluation.
The AI ​​model running the tool has passed "clinical validation", and the tool has been classified as a first-class medical device in the European Union.
The Google Health team has also built an AI model that takes into account factors such as age, gender, ethnicity and skin types - from fair-to-brown to brown-skinned.
"We developed our model and provided it with unspecified data that includes about 65,000 pictures and case data of skin diseases diagnosed during the trials, millions of coordinated images related to skin concerns and thousands of examples of healthy skin in all different population groups," Liu and Bowie concluded.
