Tibetan Medication Decision Support:Method And System Implement

Tibetan medicine is an important part of Traditional Chinese Medicine(TCM).However,making the diagnosis of the disease and therapy process depend heavily on the personal experience of doctors in the field of Tibetan medicine.At the same time,information construction of Tibetan hospital is relatively backward,and construction of a decision support system based on electronic medical records(EMR)is still in its infancy.These problems have hindered the inheritance and development of Tibetan medicine at this stage.This thesis launches the research on Tibetan medication decision support relying on the National Natural Science Fund Project of Qinghai Province — "Research on the Key Technologies of Tibetan Medicine Decision Support System Based on Data Mining".The research is carried out by combining data mining technology,recommendation algorithm and Tibetan medicine theory and using the electronic medical record data provided by Qinghai Provincial Tibetan Hospital.The main contents of this thesis includes:(1)In order to solve the problem that the classic association rule algorithm Apriori produces a large number of invalid rules when mining the Tibetan medicine pattern,a Tibetan medicine pattern mining algorithm based on the symptom-medicine pair constraint is proposed—Partition＿Apriori.This algorithm can effectively mine the Tibetan medication patterns from EMR data to guide the doctor’s medication decision.Experiments proved that the algorithm can generate fewer frequent itemsets in the process of Tibetan medical medicine pattern mining,and avoid the generation of invalid rules.(2)Analyzed the application prospects of classic recommendation algorithms in the scene of decision support for Tibetan medicines,and designed two recommendation algorithms for Tibetan medication.Ⅰ.In order to solve the user cold start problem in Tibetan medication decision support scenes,a Tibetan medication recommendation algorithm based on symptommedication pairs and collaborative filtering is designed;Ⅱ.Combining the theory of syndrome types of Tibetan medicine and fuzzy theory,a Tibetan medicine recommendation algorithm based on the mixed similarity of patient symptoms and syndrome types is designed.Experiments showed that these two Tibetan medication recommendation algorithms can generate initial prescription based on the patient’s Tibetan medicine diagnostic information,which can effectively assist Tibetan doctors in clinical medication.(3)Designed and implemented a prototype decision support system for Tibetan medication.The main function of the prototype system is to assist Tibetan doctors in making clinical medication decisions.The prototype system laid the foundation for the development of the Tibetan medicine decision support system.