AI System Based on Traditional Chinese Tongue
Diagnosis Shows High Accuracy in Disease Detection
ADELAIDE, Australia — Oct. 8, 2025 —
A research team in Australia has developed an artificial intelligence system capable of detecting signs of disease through color analysis of the human tongue, a diagnostic method traditionally used in Chinese medicine.
The system, created at the University of South Australia and tested across 60 clinical cases, achieved a 96.6% accuracy rate in identifying conditions such as anemia, diabetes, and inflammation linked to COVID-19. The tool uses a specialized imaging kiosk that captures high-quality tongue images under standardized lighting conditions to reduce variability.
Inspired by Traditional Chinese Medicine (TCM), where tongue color and coating are used to infer internal imbalances, the AI combines classical principles with modern deep learning techniques. Lead researcher Professor Javaan Chahl emphasized that the technology is not intended to replace full medical assessments but could serve as a supportive tool in preventive health screening.
The diagnostic system has been integrated into a wellness app called BenCao, which provides users with dietary and lifestyle suggestions based on tongue analysis. While the app avoids offering formal medical diagnoses, it represents one of the first commercial applications of AI trained on TCM diagnostics.
Further studies are planned to validate the tool on larger populations and a wider range of diseases.
