Research On Prevention And Treatment Of Type 2 Diabetes Mellitus Based On Data Mining

Data mining method is applied as the basic means of this article which focuses on medical data mining to the prevention and treatment of type 2 diabetes mellitus as well, so that to search out the high risks of type 2 diabetes mellitus and their level. Then gray relational method is presented to assist the injection of insulin to cure type 2 diabetes mellitus to avoid the damage of inadequacy injection of insulin.1. Basic data mining means and methods, relational rules and Apriori algorithm which is one of the most widely-applied of relational rule method are researched in this paper. Their efficiency is analyzed as well, to raise a new mind of improvement.2. Apriori algorithm, Frequent pattern growth algorithm and improved Apriori algorithm are simulated threw Matlab7.0 to analyze the applied range. It shows in the result of simulate that Frequent pattern growth algorithm fits to search long frequent items, and has a better efficiency with small support; Apriori algorithm fits to search short frequent items and has a better efficiency with large support; improved Apriori algorithm can be applied in wider fields for it has a good efficiency in both small and big support level.3. To find out the risks of type 2 diabetes mellitus and their level, 6 years’ data from the information department of one top three hospital in Lanzhou are preprocessed in a improved way, in which the preprocessed Boole data can be recognized by computer. Then input them into the module of relational rules in SPSS Clementine 12.0 to count the level of 34 attributes to programme a application to forecast type 2 diabetes mellitus for civilians first time.4. Two-step method is raised up to correctly get the 8 key points data from dynamical glucometer, and input the data from 30 data from one point in 30 minutes into the module of Two-step SPSS Clementine 12.0 to avoid the influence of dynamical saltation.5. Gray relational method is presented to analyze the 8 key points data from Two-step method, so that one most similar blood glucose level which is compared with one most standard data can be used to help injection of insulin. In this way, the damage of inadequacy infection can be avoided when the patients are without the guidance of doctors first time.