Tongue color may predict blood-biochemical data

• Published in: Artificial life and robotics Vol. 25; no. 3; pp. 383 - 387
• Main Authors: Ogawa-Ochiai, Keiko, Yamamoto, Satoshi, Namiki, Takao, Tsumura, Norimichi
• Format: Journal Article
• Language: English
• Published: Tokyo Springer Japan 01.08.2020; Springer Nature B.V
• Subjects: Artificial Intelligence; Biochemistry; Blood; Color; Computation by Abstract Devices; Computer Science; Control; Correlation analysis; Creatinine; Diagnosis; Diagnostic systems; Empirical analysis; Hyperspectral imaging; Image acquisition; Mechatronics; Medical imaging; Medicine; Original Article; Principal components analysis; Robotics; Spectra; Tongue; Kampo medicine; Hyperspectral image; Medical diagnosis; Tongue color analysis
• ISSN: 1433-5298; 1614-7456
• DOI: 10.1007/s10015-020-00620-8

Tongue diagnosis is an important diagnostic method in Japanese-traditional (Kampo) medicine, and the tongue is considered as a mirror reflecting the internal state of the body. Imbalance of water, heat, bloodstream, and to diagnose blood stagnation, blood deficiency, yin deficiency and so on in Kampo medicine can be diagnosed with tongue diagnosis. In this article, principal component analysis (PCA) of the hyperspectral images of tongues and analysis of correlation between its color and values of blood-biochemical data were performed to turn the empirical observation into the change in substantial factor. One hundred hyperspectral tongue images and blood samples were acquired from 50 outpatients, and the uncoated tongue region was segmented automatically. The spectral information of the uncoated tongue area was analyzed by principal component analysis (PCA). The component vector most representative of each blood-biochemical data was found by rotating the vector on a plane spanned by two arbitrary principal component vectors. Spectra of the tongue area without coating were extracted, the principal component analysis was performed, and correlation to each item of blood-chemistry was calculated with the vector rotation method. Creatinine, PDW, Chloride, MPV, FT4, RBC, γ-GTP, FBS, and AST showed large correlation to the tongue color spectra. We could determine correlations between tongue color spectra and items of blood-biochemistry. Our method would help non-invasive screening, and this tool should help non-trained people identify health status using concept of Kampo medicine.