AI spots diseases with 98% accuracy via tongue color
Traditional Chinese medicine has been checking tongues for over 2,000 years to find signs of disease. Nowadays, it has inspired an artificial intelligence system that can perform this function with 98% accuracy.
Ali Al-Naji, senior author and MTU Adjunct associate professor, created this wondrous device with Middle Technical University and University of South Australia researchers.
READ: AI tongue lets robots taste food
Article continues after this advertisementThe machine enables real-time diagnosis that can reduce patient wait times in hospitals. Still, human professionals will validate the AI program’s diagnoses.
How does the AI spot illnesses via tongue color?
AI predicts diseases with 98% accuracy in real-time using tongue colorhttps://t.co/IUYNuDP7I6
— Interesting Engineering (@IntEngineering) August 14, 2024
Interesting Engineering shared more details from Ali Al-Naji regarding this medical AI program. “The color, shape and thickness of the tongue can reveal a litany of health conditions,” he stated.
“Typically, people with diabetes have a yellow tongue. Cancer patients have a purple tongue with a thick greasy coating, and acute stroke patients present with an unusually shaped red tongue.”
Article continues after this advertisementAl-Naji also stated that the white tongue can indicate anemia and people with severe COVID-19 symptoms typically have deep red tongues.
That is why experts trained computer vision systems by using 5260 images categorized into tongue colors red, yellow, green, blue, gray, white and pink.
Moreover, they used six machine-learning algorithms to train the system to predict tongue color in any lighting conditions:
- Naïve Bayes (NB)
- Support vector machine (SVM)
- K-nearest neighbors (KNN)
- Decision trees (DTs)
- Random forest (RF)
- Extreme Gradient Boost (XGBoost)
Then, the researchers tested the system by making it examine 60 patient tongue images. As a result, the AI model matched tongue colors in most patient cases.
More importantly, the test proved the AI program can advance the field of medicine by spotting diseases faster. Also, Javaan Chahl, a co-author from UniSA, said they want the AI model to work on smartphones.
“These results confirm that computerized tongue analysis is a secure, efficient, user-friendly and affordable method for disease screening that backs up modern methods with a centuries-old practice,” Chahl stated.
