Research On The Combinational Soft Computing For Syndrome Differentiation In The Traditional Chinese Medicine

Syndrome differentiation in the Traditional Chinese Medicine is an important part in the theory of Traditional Chinese Medicine (TCM). It is a necessary treatment in clinical diagnosis. It is researched and discussed by those famous doctors, who lived in ancient society or live in modern society.TCM considers the body of man is an inseparable whole, and the physiological function and pathological of the body can reflect the natural changes. While the man falls ill, TCM can learn the physiological function and pathological changes by the Four Diagnoses in TCM, and treat the disease based on them. The syndrome is the base of the recipe, and the curative effect is influenced by the accurate of the syndrome different directly.The beauty of Syndrome differentiation lays in the few diagnostic techniques that accord with the most promising direction in the 21st century: no pain and no injury. However, there is a complex disorder full of contradictions, the Four Diagnoses instantly clarifies the main pathological process. Therefore, it is of great value in both clinic applications and self-diagnosis.However, Traditional Four Diagnoses have inevitable limitations that impede the TCM's applications. The analysis of the differentiation of symptoms and signs is short of the clear logic process. These disadvantages seriously block the more developing of the syndrome differentiation. Moreover, some people doubt the credibility of the TCM. Therefore, it is necessary to build a research on the objectivity and calculability for Syndrome Diagnosis in TCM, which can provide the important theory to the intelligence of the Syndrome Diagnosis and the modernization of the teaching and the scientific research.In this dissertation, several key technologies of intelligence computing for Eight Principal Syndromes and Zangfu Syndromes in Traditional Chinese Medicine are studied. Firstly the background and the current situation are expounded. Then the significance of this dissertation is elucidated. And some neuron models are derived from the concepts of the Artifice Neural Network (ANN), their features and the approximation ability to function are also analyzed. Secondly the designing model of the Genetic algorithm and the concept and the features of the Immunity Genetic algorithm are described. At last, the features and the significances of the template matching are derived from its concept. Those provide the rationales to the research of the intelligence computing for syndrome differentiation in TCM.Because the computer technologys in common used can not accord with the characters of the Diagnosis of TCM, we design an ANN model according the features of the Eight Principal Syndromes and the traditional ANN model, which is mostly used in Eight Principal Syndromes (EPSNN). The EPSNN model mainly includes three parts: the first part is the input fore-processing layer; the second part is the computing hidden layer for syndrome differencing; the last part is the output after-processing layer. The EPSNN has the memorial ability to the sample cases, and can different the syndrome more precisely.And analyzing the shortcoming of the EPSNN, we base on the theory of the Meta-population in the biology, and design a kind of Immunity Genetic algorithm—MIGA, which used to optimize the EPSNN. The MIGA can make the EPSNN have better adaptability and decreasing the request to the sample cases.On the other hand, based on the research to the specification type of syndrome and an evocation of gene targeting in biology, we offer a novel method for Zangfu Syndromes differentitation. The main thoughts are making the symptoms about the Zangfu syndromes, which include the main symptoms, the minor symptoms and the symptoms about tongue and pulse, to a model target, and making the symptoms of a clinic case to a symptoms genome about the clinic case, which is used to shoot at a syndrome model target. The experiment results about Syndromes differentiation for the apoplexy and the chronic gastritis approve that the method is effective. Our method proposed provides a researching method in the objectivity and calculability of Zangfu Syndromes differentiation, and also have some clinical value of diagnosis and application.The main content of this dissertation is satisfactorily performed by the study results in the paper and the conclusion shows the study results are correct and feasible. Therefore, this research has certain value in the field of Information Process for Traditional Chinese Medicine and has important practical significance.