Researchers have developed a novel computer algorithm that can predict various diseases like diabetes or stroke with 98 percent accuracy by analysing the colour of the human tongue.
The imaging system developed by Middle Technical University (MTU) and the University of South Australia (UniSA) in Australia can diagnose diabetes, stroke, anaemia, asthma, liver and gallbladder problems, COVID-19 and other vascular and gastrointestinal diseases.
“The colour, shape and thickness of the tongue can detect a number of health conditions,” said Ali Al-Naji, adjunct associate professor at MTU and UniSA.
He said, “Typically diabetics have a yellow tongue; cancer patients have a purple tongue with a thick greasy coating; and patients who suffer an acute stroke have an abnormally red tongue.”
This breakthrough was achieved through a series of experiments using 5,260 images to train a machine learning algorithm to detect tongue colour.
The researchers received 60 tongue photos from two teaching hospitals in the Middle East, depicting patients with various health conditions. The AI model matched the tongue color to the correct disease in almost all cases.
The paper, published in Technologies, describes how the system analyses the colour of the tongue to provide a diagnosis in real-time, showing that AI could significantly advance medical practices.
Al-Naji explained that the AI is mimicking a 2,000-year-old technique from traditional Chinese medicine, which uses the colour, shape and thickness of the tongue to diagnose health problems.
For example, people with diabetes usually have a pale tongue, while cancer patients have a purple tongue with a thick greasy coating. Stroke patients often have an unusually red tongue. A white tongue can be a sign of anemia, severe COVID-19 cases have a dark red tongue, and a blue or purple tongue may indicate vascular or gastrointestinal problems or asthma.
The study placed cameras at a distance of 20 centimeters from the patient to measure tongue color, and the imaging system predicted health conditions in real time.
Co-author UniSA Professor Javan Chahal said the technology could eventually be adapted for use with smartphones, making disease screening more accessible.
(Except for the headline, this story has not been edited by NDTV staff and is published from a syndicated feed.)
