文献类型：期刊文献

英文题名：Study on intelligent syndrome differentiation in traditional Chinese medicine based on information fusion technology

作者：Wang, Yiqin[1]; Guo, Rui[2]; Yan, Haixia[1]; Li, Fufeng[3]; Xia, Chunming[4]; Xu, Zhaoxia[1]; Xu, Jin[1]

机构：[1] School of Basic Medicine, Shanghai University of TCM, Shanghai, China; [2] Center for Information of TCM, Shanghai University of TCM, Shanghai, China; [3] Center for Teaching Experience, Shanghai University of TCM, Shanghai, China; [4] Center for Mechatronics Engineering, East China University of Science and Technology, Shanghai, China

年份：2010

起止页码：698

外文期刊名：2010 IEEE International Conference on Bioinformatics and Biomedicine Workshops, BIBMW 2010

收录：EI(收录号：20110913708930)

语种：英文

外文关键词：Neural networks - Support vector machines - Artificial heart - Information fusion - Diagnosis - Diseases

摘要：Objective: Establish four-diagnosis syndrome differentiation model of Traditional Chinese Medicine (TCM) based on information fusion technology (four diagnostic methods refer to inspection, auscultation and olfaction, inquiry and pulse-taking. Method: Apply the objective detection instruments of four-diagnostic method to collect four-diagnosis objective information of 509 cases of clinical heart-system patients, then adopt multiple artificial neural network of single output and multiple- support vector machine to establish recognition model of syndrome above. Result: Recognition rates of the 6 syndromes, Deficiency of Heart Qi, Deficiency of Heart Yang, Deficiency of Heart Yin ,Phlegm, blood stasis, Stagnation of Qi, by multiple artificial neural network of single output, are respectively 60.67%inverted commas 78.08%inverted commas 65.16%inverted commas 60.11%inverted commas 62.35% and 87.07%, whereas, by multi-class support vector machine, respectively 73.20%inverted commas 81.70%inverted commas 68.63%inverted commas 50.33%inverted commas 76.47%inverted commas 85.62%. Conclusion: TCM four-diagnosis syndrome differentiation model set up based on SVM is of high quality with compare with artificial neural network. ?2010 IEEE.