Research On Pragmatizing The Tongue Image Analysis Instrument Of Traditional Chinese Medicine

Tongue diagnosis is an important section of looking diagnosis of traditional Chinese medicine. It one of the most important clinical examining manner. Information about the state of the internal organs can be obtained by observing tongue body and tongue-coating. So far, Chinese clinicians have been making tongue diagnoses by means of subjective visual methods which is subject to large inter- and intra-observer variation and lacks objective criterion. Moreover, it difficult to manage the tongue information efficiently. However, those disadvantages can be improved with the development and pragmatizing of Tongue Image Analysis Instrument of Traditional Chinese Medicine (hAl) which took advantage of the technique of modern signal and information processing and integrated the clinical experience of the traditional Chinese medicine experts. TIAI can acquire digital tongue image, analyze the tongue image and provide more objective quantitative analysis results for doctors. Moreover, it can manage the tongue information efficiently by Tongue Image Database. So TIAI plays an important role in clinical diagnosis, science research and teaching of traditional Chinese medicine.The major goal of this thesis is to research the pragmatizing of TIAI. The author had assisted Ph.D. candidates in researching some key technique of TIAI. However, it is still far from the practical application. Therefore, the author has pragmatized TIAI independently. The main achievements are as follows.Based on analyzing the operation flow of tongue diagnosis, the main control software is designed and developed.The factors that may influence the color reproduction are analyzed and an online color calibration method based on the combination of color evaluation and colorimetric matching is developed. It is capable of rendering images captured under different conditions looking similar to eyes.A dual snakes model based approach of image segmentation is proposed for extracting tongue areas from the tongue images. The tongue contour is defined as a Catmull-Rom spline-based snake. An automatic segmentation algorithm and a semi-automatic algorithm with preferable robustness based on the dual snakes model are implemented.A tongue image database is designed and developed for managing the information of tongues. And a Chinese typical tongue image database was found which consists of 50 kinds, more than 1000 tongue images and their information.The debugging of the second TIAI is fulfilled and its pragmatizing is achieved. It works in Beijing Traditional Chinese Medicine Hospital now.The pragmatizing of the TIAI provides the necessary technical support for the development of traditional Chinese medicine tongue diagnosis.