For centuries, doctors have peered into patients' mouths, examining the tongue for tell-tale signs of illness. Now, a revolutionary leap forward sees artificial intelligence (AI) taking on this ancient practice, analysing the tongue's appearance with startling precision to flag early warnings for conditions like diabetes and stomach cancer.
How AI is Learning the Language of the Tongue
A comprehensive review of more than 20 scientific studies, reported in the journal Chinese Medicine, concludes that AI programs have become so adept at spotting disease indicators that hospitals could soon deploy them to aid diagnosis. These systems are trained on vast databases containing thousands of photographs of tongues from both sick and healthy individuals.
"AI learns by identifying statistical patterns in large collections of tongue images paired with the patient's clinical or health-related data," explains Professor Dong Xu, a bioinformatics expert at the University of Missouri. The technology scans for minute visual characteristics—such as colour distribution, surface texture, moisture, and swelling—that occur more frequently in people with specific ailments.
One standout 2024 study, published in Technologies, demonstrated remarkable accuracy. The AI program correctly identified 58 out of 60 patients with diabetes and anaemia based solely on a picture of their tongue.
From Gastric Cancer to Glandular Fever: What Your Tongue Reveals
The principle is far from new. "The tongue is referred to as the mirror of general health," says Professor Saman Warnakulasuriya, an emeritus professor of oral medicine at King's College London. Traditional medicine has long linked specific tongue changes to underlying issues.
For instance, a smooth, shiny tongue can signal anaemia due to deficiencies in iron or B vitamins, which are essential for cell renewal. A dry tongue may be an early diabetes symptom, while a thick white coating can indicate infection. Subtle colour and texture changes, like patchy colour loss or areas of redness, have been linked to inflammation in the digestive tract.
AI's power lies in its ability to detect patterns that might elude the human eye. A 2023 study in eClinicalMedicine found AI could distinguish patients with gastric cancer from healthy volunteers with 85 to 90 per cent accuracy—a rate comparable to standard diagnostic tests like gastroscopies or CT scans.
A Powerful Assistant, Not a Replacement
Despite its promise, experts unanimously stress that AI tongue analysis is a tool for clinicians, not a standalone diagnostician. The technology has limitations; it may misinterpret a pale tongue caused by poor circulation as anaemia, for example, because it works on statistical correlations without understanding cause.
Professor Bernhard Kainz of Imperial College London suggests the technology is most reliable as a broad health checker. "Used appropriately, AI tongue analysis can help prioritise care and reduce missed early signs, but it should complement, not replace, established diagnostic pathways and clinical judgment," he advises.
Professor Warnakulasuriya agrees, noting that while AI can give doctors confidence to narrow down a diagnosis, "it is always necessary to confirm the diagnosis by conducting appropriate laboratory tests." Factors like lighting, camera quality, diet, and medication can also affect the tongue's appearance, potentially confusing the AI.
As this technology continues to evolve, its integration into UK healthcare could mark a significant step towards faster, non-invasive preliminary screenings, helping GPs decide who needs urgent further investigation.
