摘要
目的:基于数据挖掘技术,探索弥漫性肺间质疾病中医证候诊断标准模型。方法:采用流行病学方法,收集弥漫性肺间质疾病临床调查资料475份。采用人工神经网络(Artificial Neural Network,ANN)、模糊系统(fuzzy system,FS),构建基于动态Kohonen网络的自适应模糊推理系统模型。通过Fisher-iris数据检验模型可靠性。最后用该模型对临床数据挖掘,依据中医基础理论,获得弥漫性肺间质疾病常见证候诊断标准,并检验其合理性。结果:弥漫性肺间质疾病常见有痰瘀阻肺证、肺肾气阴两虚证、肺肾气虚证、痰热壅肺证、痰湿壅肺证等5个证候。证候诊断标准诊断符合率为73.8%。结论:该模型适用于中医证候诊断标准的研究,方法合理,结果可靠。
Objective：To investigate the model of syndrome diagnostic criteria of diffuse interstitial lung disease（DILD）. Methods：Based on method of epidemiological investigation ,475 patients＇ data of DILD collected. Artificial neural network （ANN） and fuzzy system （FS） was used to build adaptive fuzzy inference system model, based on dynamic Kohonen network. And test the reliability of the model by Fisher - iris. At last, the model was used for clinical data mining, based on basic theories of Chinese medicine, access to common syndromes diagnostic criteria of DILD, and legitimacy tested. Results ： There are five common syndromes of DILD,such as syndrome of phlegm and blood stasis obstructing lung,syndrome of deficiency of both qi and yin of lung and kidney, syndrome of lung - kidney qi deficiency, syndrome of phlegm - heat blocking pulmonary and syndrome of phlegm - dampness obstruct lung. Syndrome diagnostic criteria diagnosis rate is 73.8%. Conclusion：The model is applicable to TCM diagnostic criteria research, its methods is reasonable, and results is reliable.
出处
《辽宁中医杂志》
CAS
北大核心
2010年第12期2333-2335,共3页
Liaoning Journal of Traditional Chinese Medicine
基金
河南省高校新世纪优秀人才支持计划(2006HANCET-05)
