| Objective:Traditional Chinese Medicine (TCM) includes a range of traditional medical practices originated from China. Examination of the condition of the tongue is one of the most valuable and widely used diagnostic methods in TCM diagnostics. In tongue diagnostics, changes in tongue color are sensitive indicators of internal pathological symptom and indispensable guideline for overall health state. However, the further development of traditional tongue diagnosis is limited by its dependence on individual visual sensation and experience. There are three problems in the field of tongue color analysis. The first one is the standardization of observation status. The second one is that TCM only offer the results of qualitative analysis, without any results of quantitative analysis. The third one is the subjectivity in the diagnostic results. In a word, this will undoubtedly limit the application of tongue color analysis in clinical practice. In recent years, many researchers are dedicated to the combination of modern pattern recognition and image processing technologies, trying to find solutions to the non-quantitative issue of traditional tongue diagnosis.In this report of computerized tongue diagnosis, we investigate tongue color analysis and diagnosis classification, including:Methods:(1) The color checker customization is introduced. This is a part of preprocess of tongue color analysis. We use the Uniform Experimental Design to select the best group of tongue color checker.(2) Along with the rapid growth of medical data, image retrieval, a kind of technology for browsing, searching and retrieving similar images of the given image, has become increasingly important from a large database of digital images. Tongue coating is the most important characteristic to reveal the pathological changes of the tongues for identifying diseases. In this paper, an efficient and effective technique is proposed to retrieve coating images. We obtain the pixel template value of pixels by applying thresholding segmentation based relative entropy. Then we use a Reduced K Nearest Neighbor algorithm to extract20-dimension feature vector based on a prior layout distribution. Finally, a distance based the cumulative ratio is proposed for tongue coating image matching.(3) Examination of the tongue condition is a standard diagnostic method in Traditional Chinese Medicine (TCM) and takes account of a wide variety of features including shape, texture, and color. The terms "warm","neutral", and "cool" are used to refer to a kind of chromatics characteristic of the tongue color and are associated with various health states. In this paper, we propose a semi-supervised scheme for tongue color analysis on "warm or cool". In its training part, the proposed scheme firstly makes use of a classical clustering algorithm, Expectation Maximization, to divide all pixels in tongue gamut into many clusters. Secondly, we construct two auxiliary images for each cluster and manual labeling endows each cluster with a category label of "warm or cool". Thirdly, each trained (or labeled) category on "warm or cool" is set up by sum some clusters approximately. Finally, in the testing part, we use a lookup table to divide all pixels in an input image into three distinct categories of "warm or cool"(4) Tongue diagnosis is a distinctive and essential diagnostic method. The color category of the tongue can be utilized to discover pathological changes on the tongues for identifying diseases. In this paper, a novel scheme is established which classify tongue images into various categories, including coating and substance categories. Firstly, we proposed a two level hierarch clustering method for quantizing all pixels into numerous vectors of feature value. Each vector can code a very small sub-class in RGB color space. Secondly, we utilized the vectors’distribution of these sub-classes to represent approximate chromatic information of tongue images. Then, a Bayesian Network is employed to model the relationship between these quantized vectors and tongue color categories.(5) Finally, we proposed a fusion system, which combine the online and offline methods of customized color checker to improve the accuracy and quality of the image color correction. Combining the idea of hierarchy clustering which we mentioned before, the accurate values of each color part in the color checker can calculated based on the customized color checker.Results:(1) In this way, the quantity of color checker is small and the design process of color checker customization is very easy.(2) The experimental results indicate that the proposed scheme eliminates the imprecision and uncertainty associated with medical tongue coating analysis.(3) In experiments conducted on a total of392tongue samples, our system achieved an accuracy of91.1%.(4) The effectiveness of this scheme is tested on a group of418tongue images, and the classification results are reported.(5) And the accuracy and quality of the image color correction can be improved by iterative algorithm. The subjective factors and calculation errors are limited at utmost by this fusion system.Conclusions:In this report, we carry out some investigations on tongue color analysis and classification, which will be helpful for the computerized tongue diagnosis. |
