AI system uses tongue colour to identify multiple health conditions
What you need to know:
- They explained that different tongue colours indicate different health conditions.
- People with diabetes often have a yellow tongue, while cancer patients may have a purple tongue with a thick, greasy coating.
A recent study has shown that a computer algorithm can predict various diseases with 98 per cent accuracy by analysing the colour of the human tongue.
The innovative imaging system, developed by researchers at Iraq's Middle Technical University and the University of South Australia, has the potential to diagnose a range of health conditions, including diabetes, stroke, anaemia, asthma, liver and gallbladder disease, Covid-19 and various vascular and gastrointestinal problems.
The results, published in the journal Technologies, show how the system uses tongue colour analysis to provide an instant diagnosis, demonstrating the potential of artificial intelligence (AI) to revolutionise medical diagnostics.
The AI model was able to correctly match the tongue colour to the corresponding condition in almost every case.
According to the researchers, the AI system essentially replicates a 2,000-year-old diagnostic technique used in traditional Chinese medicine that examines the colour, shape and thickness of the tongue to detect health problems.
They explained that different tongue colours indicate different health conditions. For example, people with diabetes often have a yellow tongue, while cancer patients may have a purple tongue with a thick, greasy coating.
Similarly, patients who have suffered an acute stroke typically have a red, unusually shaped tongue, while a white tongue can be a sign of anaemia and a deep red tongue has been linked to severe cases of Covid-19.
The researchers trained machine learning algorithms using 5,260 tongue images to detect subtle colour differences associated with different diseases.
The images were collected from patients with various health conditions at two teaching hospitals in the Middle East.
In the study, tongue colour was captured using cameras positioned 20 centimetres from the patients' tongues. The imaging system was able to predict the patient's health status in real-time, demonstrating the potential of this technology for rapid and non-invasive diagnostics.
The researchers suggest that this technology could eventually be integrated into smartphones, allowing people to diagnose diseases using their mobile devices.
As the technology develops, it could become an important tool both in clinical settings and for personal health monitoring, potentially improving early detection and treatment of various diseases.
"These results confirm that computerized tongue analysis is a secure, efficient, user-friendly, and affordable method for disease screening that supports modern diagnostic methods with an ancient practice," said Prof. Javaan Chahl, co-author of the study.
