Google has previewed a new tool powered by artificial intelligence technology that it says could help identify skin problems . The tool, announced at the company’s Google IO conference last week, uses artificial intelligence to identify potential skin conditions. However, it is not intended to replace medical advice, the tech giant said.
A database with thousands of photos
To use it, patients must upload three photos of a problem area to the tool, which Google says can recognize up to 288 different skin, nail and hair conditions. According to an article published by Google researchers in the journal Nature Medicine just a few weeks ago, the tool was developed using a set of around 65,000 anonymous images and case data for diagnosed diseases, taken from a total of 16,114 individual cases.
However, some experts assure that the images used in this study are not completely representative of all skin tones, which dermatologists classify according to the Fitzpatrick scale, with Type 1 skin being the palest and Type 6 the most. dark
Unrepresentative and non-diverse images
The company has recognized that the skin tones with which the application works are not completely representative, but they assure that they are working to include all possible varieties and make the tool much more effective.
The lack of examples of darker skin types in the trial version of the tool could negatively affect patients with darker skin tones who use it. The experts most critical of this application assure that there is a lot of evidence showing that persistent racial and gender inequality in society is reflected in skewed artificial intelligence data sets, which translates into algorithmic bias and leads to negative results in people. structurally marginalized.
In a statement, a Google spokesperson defended itself, saying that the data presented in the Nature Medicine study was out of date. “Our work is the culmination of more than three years of continuous development. After our initial research in Nature Medicine and JAMA Network Open, we have continued to refine the technology on which our tool is based, including obtaining additional data sets [ …] Equity has, and will continue to be, a key focus and this includes working with data sets that include different ethnicities, skin types and age groups, and partnering with clinicians and experts who have experience working with communities of color “, the spokesman sentenced.
However, a later study published by Google researcher Dr. Yun Liu in April recognized the lack of diversity in the images used. The tool, which is currently not available to the general public, has been approved for use in the EU as a Category I medical tool, the lowest risk category that encompasses other non-invasive tools such as stethoscopes.
