ADELAIDE, Australia — A practice rooted in Chinese herbal medicine for over two millennia — examining the human tongue for signs of illness — has found itself at the forefront of modern health care, thanks to artificial intelligence (AI).
Tongue diagnostic systems, traditionally employed by Chinese herbalists, are gaining popularity worldwide, particularly in the context of remote health monitoring. A study conducted by researchers from the Middle Technical University (MTU) in Baghdad and the University of South Australia (UniSA) now provides further evidence of the remarkable accuracy of this technology in detecting diseases.
Engineers from MTU and UniSA harnessed the power of a USB web camera and computer to capture images of patients’ tongues, focusing on 50 individuals afflicted with conditions like diabetes, renal failure, and anemia. They then compared the captured tongue colors with a comprehensive database of 9,000 tongue images.
Leveraging advanced image processing techniques, researchers achieved a diagnostic accuracy rate of 94 percent, as compared to traditional laboratory results. The system automatically generated voicemails specifying the patient’s tongue color and the detected disease, which were sent as text messages to either the patient or their designated healthcare provider.