Developing a tongue coating index for Traditional Chinese Medicine
Choy, Meng Xuan
Date of Issue2017-05-23
School of Mechanical and Aerospace Engineering
Tongue coating analysis is part of tongue diagnosis in Traditional Chinese Medicine (TCM). Besides denoting the nature and site of the pathogenic factors, the tongue coating is believed to reflect the constitution of the stomach. [1, 11-13] To date, tongue diagnosis has always been reliant on the visual judgment (that is conducted in an environment where factors such lighting conditions are not kept constant) and experience of TCM practitioners. Hence diagnosis is inconsistent. There is a lack of scientific approach towards TCM diagnosis, although there has been research effort to quantify the tongue coating in terms of its color, thickness, moisture and texture. [5,8,9] In this Final Year Project (FYP), the author will be focusing on establishing a Tongue Coating Index (TCI) that quantifies the thickness of the tongue coating through the method of discerning color differences between 2 reference points and the 5 tongue regions. The tongue coating could be quantified by both its color and thickness. The author proposes to quantify the tongue coating thickness through the development of an index based on the concept of color differences. The objective of this FYP is to formulate an Index for tongue coating. The scope of work will involve analysis of a large number of previously acquired tongue images. The difference between two tongue colors, represented in the 3-dimensional color space of L-a-b, can be defined as: d=√(〖(x2-x1)〗^2+〖〖(y2-y_1)〗^2+〖(z2-z1)〗^2 ) where d represents the distance (Index) between the 2 tongue colors. (x1, y1, z1) is the coordinates of the reference color point (L-a-b)1, and (x2, y2, z2) is the coordinates of tongue color for analysis (L-a-b)2 Two methods of TCI were investigated. Approach 1, where L-a-b values of each tongue region were compared with those of the tongue tip, and Approach 2, where L-a-b values of each tongue region were compared with those of the ideal tongue color (Pink). Results obtained show that Approach 1 was more consistent with visual perception of post-corrected tongue images in the database. Besides the establishment of the TCI, the author has also strengthened the old database with a significant amount of crucial tongue information obtained through data mining, making it a more effective and efficient system for tongue diagnosis. With the updated database system, any information concerning a subject’s tongue and health condition can be easily retrieved.
Final Year Project (FYP)
Nanyang Technological University