A team of researchers from the University of South Australia and Iraq’s Middle Technical University has developed an artificial intelligence (AI) model that can accurately diagnose certain illnesses by analyzing a patient’s tongue. The novel technology draws inspiration from traditional Chinese medicine, which has been using the tongue as a diagnostic tool for over 2,000 years.
The AI model is trained to identify unique characteristics and features on the tongue, such as color, shape, and thickness, that can reveal possible health issues, including cancer, diabetes, asthma, and gastrointestinal problems. The researchers found that tongue color is particularly important in detecting illnesses and monitoring their progress.
According to the study’s senior author, Ali Al-Naji, certain scenarios can be identified by analyzing the tongue. For example, people with diabetes typically have a yellow tongue, while those with cancer may have a purple tongue with a thick greasy coating. Acute stroke patients may present with an unusually shaped red tongue, and a white tongue can indicate anemia.
The AI model was trained using two data sets, including 5,260 images of different tongue colors across various light conditions and saturations. The system achieved a remarkable precision rate of over 98% in detecting ailments based on tongue color changes. In a test group of 60 tongue images, the program achieved an accuracy rate of 96.6%.
The researchers believe that their AI model could eventually be used to provide a secure, efficient, and cost-effective method for disease screening in medical facilities.
Source: https://www.popsci.com/technology/ai-tongue-scanner/