Chinese traditional medicine is a kind of precious deposits. The analysis of somefamous herbalist doctor's experience can not only enrich the systEM of Chinesetranditional medicine, but also greatly improve it. So it is a vital task for our Chinese tolearn and master the experience. An old doctor of Chinese traditional medicineaccumulated the massivevaluable experience in his clinical practice, but these experiencesis hided in the massive clinical medicalrecords, solely depends on person's manualcollection is insufficient. So we must have to use computer technology which is aformidable weapon. How to use the computer technology in the massive medical records,which is a big key to inherit the traditional Chinese medicine experience of the herbalistdoctor.This article takes the old doctor of traditional Chinese medicine chronic gastritisclinical diagnosis medical records as the information.First we use the factor analysis toanalyze dialectical and the symptom relations from the whole, then select the appropriateillness and the factor, and obtain the relations chart of the chronic gastritis and thesymptom.After that we use the EM algorithm to study and to improve the hidden thestructural model and the related parameters We obtain a more accurate relations chart of thechronic gastritis and the symptom.Finally according to the relations hart which isobtained,we use Bayesian network to diagnose the symptom, then we achieve the finalgoal.Our final diagnosis result, basically conforms to the knowledge of traditional Chinesedoctor.The obtained dialectical and the symptom relations chart has also provided thereference to the Chinese medicine diagnosis. |