Algorithms of artificial intelligence (AI) have long been applied for medical purposes, including – to diagnose various diseases. New Google developments are sets of software tools to determine the problems with skin and tuberculosis.
Using Google technology, skin diseases can be diagnosed literally using a smartphone camera. The company claims that certain problems of this nature have billions of people, but due to the lack of specialists, hardware and software diagnostic complexes become indispensable.
The user can simply make three snapshot of the problem area of the skin from various angles, answer a few questions – on the basis of the data obtained, the maximally suzzings the circle from hundreds of possible diagnoses for further consideration of the analysis results by specialists.
According to Google, artificial intelligence was trained on the basis of millions of pictures relating to different demographic groups, but until the technology only has an auxiliary value. The development was considered in the Office for Food Control and Drug Administration, but it cannot serve a full-fledged visit to the professional visit. Pilot “run-up” tool in the field of Google intends to spend later this year.
For the diagnosis of tuberculosis and the subsequent direction of potential patients for additional examinations, Google used systems of deep learning. AI analyzes the chest X-rays before a person will be directed to passing more expensive special tests.
According to Google, the technology developed by the company gives no more false-positive and false-negative results than 14 of the participating radiologists. To create an artificial intelligence system, Google applied “impersonal” data from nine countries, and then checked the trained “intelligence” on patients from five countries. If the doctors do not consider the results of Google’s evaluation by default, they can solve independently, send patients to additional examinations or not.
Here you can download a full report on technology. In the future, the Company intends to expand the study at the expense of partnership with hospitals in India and Zambia.