| Type 2 diabetes is one of the main diseases which threaten people's health. Statistics show that there are 20million patients with type 2 diabetes in China, the number of patient is NO.2 in the world. Traditional Chinese Medicine (TCM) plays an important role in improving clinical symptoms and signs, patient's quality of life and preventing the chronic complications. As the diagnosis of TCM syndromes and the evaluation of the clinical therapeutic efficiency at present, there is not a common, regular and objective diagnosis standard system, this current situation makes the clinical curative effect assessment not provide enough evidences. Therefore, it is important to study the normalization and standardization of the syndromes of type 2 diabetes. According to the study mode of the combination disease with differentiation of their syndromes, this research adopts clinical epidemiology and health statistics to study diagnosis standard of TCM syndrome differentiation of type 2 diabetes based on computational intelligence (CI).Objective: To study diagnosis standard of Traditional Chinese Medicine (TCM) syndrome differentiation of type 2 diabetes based on computational intelligence (CI) and statistical methods, then try to explore the methods and approaches for establishing diagnosis standard of TCM syndrome differentiation.Methods: Firstly, the Fisher-Iris data was selected by internet. Secondly, the document data from 1978 to 2004 on type 2 diabetes syndromes were gathered by way of document database index by internet. Syndromes of patients with type 2 diabetes from three hospitals in 2004～2006 were investigated by launching the clinical questionnaires. The databases of document and clinic were set up by data pre-preprocessed.The distribution regularity of syndromes about type 2 diabetes was explore and summarized by the methods of logistic regression, clustering analysis and factor analysis, and the Artificial Neural Network (ANN) and fuzzy system (FS) were used to study on model establishment and application of syndromes diagnosis standard for the type 2 diabetes.Results: 233 documents and 1444 cases were used for analysis. The results of this study included several aspects.1 Based on the statistical methods:①The results indicated that 26 syndrome factors were found out in document study, which included that kidney, lung, spleen and stomach were the main location syndrome factors, and Qi deficiency, Yin deficiency, fire and heat, Yang deficiency and blood stasis were the main nature syndrome factors. In clinical research, 15 syndrome factors were found out, and the highest frequency of location syndrome factor was spleen (35.9%), stomach, kidney, and lung were higher. The frequency of Qi deficiency was the highest (54.1%) in the nature syndrome factor, Yin deficiency, Qi stasis and blood stasis were higher.②In the document study, there were seven patterns in the combinative model of syndrome factors, single syndrome, the combination of two syndrome factors and the combination of three syndrome factors were the main patterns, the cumulative percentage was 91.1%. In clinical research, there were eight patterns in the combinative model of syndrome factors, from the combination of two syndrome factors to the combination of six syndrome factors were the main patterns, the cumulative percentage was 87.5%.③The results showed that among 216 symptoms appeared in the 175 documents, the highest frequency of symptom was hyperdiuresis, the frequency of red tongue, tired and debilitation, thirsty and polydipsia, overeating and limosis, dry throat and dry tongue were higher; The results of the clinical research indicated that several symptoms which frequency was more than 50% or equal to 50% and the average scores more than 4 or equal to 4, were tired, debilitation, thirsty and polydipsia, upset, memory descent, tantrum, body felling heavy, illness deteriorated because of anger, illness deteriorated because of fatigue. These symptoms could be the main symptoms of the type 2 diabetes.④The outcome of the document study showed that Qi-Yin deficiency, Yin-Yang deficiency, heat produced due to Yin deficiency, blood stasis, deficiency of fluid due to dryness and heat were the main syndromes of type 2 diabetes, meanwhile, the main symptoms and the sub-symptoms of the six syndromes were screened out by adopting statistical description, logistic regression and clustering analysis.⑤In the clinical research, 21 common factors were extracted by exploratory factor analysis, the cumulative contribution rate was 60.257%, and 13 groups were clustered in these factors by clustering analysis.2 Based on computational intelligence:①Based on the Fisher-Iris data, when the dynamic layer was turn into stability, the never cells of the dynamic layer were added up to 9. 3 fuzzy rules were acquired and the rate of coincidence identity of test sample is 94%.②The results gained based on document data showed that the never cells of the dynamic layer were added up to 21, 9 fuzzy rules were acquired and the rate of coincidence identity of test sample was 80%. The main symptom and the sub-symptom of six main syndromes of type 2 diabetes were selected according to the transformed rule, the six syndromes were Qi-Yin deficiency, Yin-Yang deficiency, heat produced due to Yin deficiency, blood stasis, deficiency of kidney-Yin, deficiency of fluid due to dry lung.③The outcome obtained based on clinical data showed that the never cells of the dynamic layer added up to 122, 27 fuzzy rules were acquired and the rate of coincidence identity of test sample was 73%. Qi-Yin deficiency, blood stasis, deficiency of fluid due to dry lung, deficiency of kidney-Yin, obstruction by dampness were selected as the main syndromes, and basic clinical features of these syndromes were mined out.3 According to the results of document study and clinical research, which instituted the diagnosis standard of the main syndromes, and these syndromes included Qi-Yin deficiency, blood stasis, deficiency of fluid due to dry lung, deficiency of kidney-Yin, and obstruction by dampness.Conclusions: 1 The main syndrome factors which affected the syndrome differentiation of type 2 diabetes included eight factors, and kidney, lung, spleen and stomach were the main location syndrome factors. Qi deficiency, Yin deficiency, fire and heat, and blood stasis were the main nature syndrome factors.2 Tired, debilitation, thirsty and polydipsia were the basic clinical symptoms of type 2 diabetes, and the symptoms of Qi deficiency appeared to be the main clinical features of type 2 diabetes.3 General syndromes diagnosis criterion of type 2 diabetes included 5 types as:①Deficiency of fluid due to dry lung consisted of main symptoms and sub-symptoms. The main symptoms included hyperdiuresis, pollakiuria, dry throat, dry tongue, thirsty and polydipsia, red tongue, yellow moss. The sub-symptoms included overeating and limosis.②Qi-Yin deficiency consisted of main symptoms and sub-symptoms. The main symptoms included tired and debilitation, palpitation, red tongue, light moss, little moss, thready and rapid pulse. The sub-symptoms included thirsty and polydipsia, dry throat and dry tongue, agrypnia and dream, upset, night sweet and emaciation③Blood stasis consisted of main symptoms and sub-symptoms. The main symptoms included limbs numb, limbs pain, petechia tongue, dark texture of tongue, choppy pulse. The sub-symptoms included wiry pulse and dark texture of face.④Obstruction by dampness consisted of main symptoms and sub-symptoms. The main symptoms included fatigue, greasy tongue. The sub-symptoms included abdominal distension, gastric distension, nausea.⑤Deficiency of kidney-Yin consisted of main symptoms and sub-symptoms. The main symptoms included hyperdiuresis, pollakiuria, dry tongue, dry throat, feeble of waist and lap, chyluria, red tongue. The sub-symptoms included heat of hand and foot, little moss.4 The methods of logistic regression, clustering analysis and factor analysis adopted in the study of diagnosis standard for syndrome differentiation of type 2 diabetes had several limitation, and the results showed that the model which was set up by use of CI methods possessed higher reliability, it could be used for the study in syndrome diagnosis standard system of type 2 diabetes.
|
PDF Full Text Request