AI Algorithm Achieves 98% Accuracy in Diagnosing Diseases Through Tongue Colour Analysis
Researchers from Iraq and Australia created a new computer program that can accurately predict 98% of diseases just by looking at a person's tongue colour. This is a huge step forward in medical diagnosis. Middle Technical University (MTU) and the University of South Australia (UniSA) worked together to create the cutting-edge image system, which is a new take on an old way of diagnosing health problems.
The study, which was released in the journal Technologies, talks about how the system can find a lot of different illnesses, like diabetes, stroke, anaemia, asthma, COVID-19, liver and gallbladder diseases, and a number of vascular and digestive problems. We trained the algorithm with 5,260 pictures of tongues, enhancing its ability to match tongue color with various health problems.
In a set of tests, the AI model was put to the test with 60 images of tongues from people in two teaching hospitals in the Middle East who had different illnesses. It was amazing how well the AI could guess what the health problem was almost every time.
Ali Al-Naji, the lead author of the study and an adjunct associate professor at both MTU and UniSA, talked about how important this achievement was. According to him, the AI system is essentially replicating a diagnosis method that traditional Chinese medicine has employed for over 2,000 years. "The colour, shape, and thickness of the tongue can reveal a litany of health conditions," he told us. Al-Naji continued, stating that individuals with diabetes typically have a yellow tongue, those with cancer typically have a purple tongue with a thick, greasy coating, and those who have experienced an acute stroke typically have a red tongue with an abnormal shape.
He also said that changes in tongue colour can be a sign of other health problems. For example, a white tongue can mean you have anaemia, a deep red tongue can mean you have serious COVID-19, and an indigo or violet tongue could mean you have problems with your blood vessels, your digestive system, or your asthma.
In real life, we placed cameras 20 centimeters away from a patient's tongue to record its colour, and AI provided real-time health predictions. This method not only works right away, but it also combines old knowledge with cutting-edge technology.
Professor Javaan Chahl of UniSA, who helped write the study, talked about how useful this technology could be in the future. "Hopefully, in the future, we can use a smartphone to diagnose medical conditions." Chahl went on to say, "These results confirm that computerised tongue analysis is a safe, effective, easy-to-use, and inexpensive way to screen for diseases that builds on a practice that dates back hundreds of years."
The study's findings suggest that this AI-powered diagnostic tool could revolutionize disease detection and tracking, providing a straightforward yet efficient solution that integrates traditional and modern medical approaches. As the technology improves, it could become an easy-to-use tool for finding diseases early on, which could help millions of people around the world.