Diabetes Analysis Model Research Based On Breath Odor

Diabetes is a kind of chronic disease which is harmful for human health. The incidence rate is rising with the lives of people are getting better. According to the record of the World Diabetes Federation, there are290million diabetes patients. There are more and more diabetes patients in developing countries such as India and China. According to the research, the number of diabetes patients has increased nearly five times during the past20years in China. The number of people died because of diabetes is the third largest of the number of people died because of unnatural fatal disease. It is only after the number of cancer patients and cardiovascular patients. Is also the main cause of blindness, kidney failure and amputation. The symptom of diabetes is not clearly in the early stage. People has diabetes did not realized they have diabetes until the conditions get more serious. So it is very important of diagnose diabetes as early as we can.The common diagnostic methods of diabetes is blood glucose testing. It is a traumatic method. There are other ways to diagnose diabetes such as plantar pressure detection. But the accuracy of this method is low. According the research, the breath exhaled by diabetes patients contains a lot of acetone. So we can detect the proportion of acetone in breath gas to diagnose diabetes. And this diagnostic method is a noninvasive examination method. This dissertation is mainly about the research of difference between diabetes and healthy. The most important part of this dissertation is the research of the condition of diabetes.This dissertation discusses breathing diagnosis of diabetes. It mainly contains pattern classification of gas exhaled by human. This dissertation discusses the preprocess of the collected sample at first. Then it shows the methods of feature extraction. There are two kinds of feature. The two kinds of feature are geometric feature and characteristic parameter of curve fitting. Then the dissertation shows the results of classification among the diseases and the healthy. The classification methods are PCA and LDA. The conditions of diabetes are not like other diseases. The treatment ways are different for different levels of the conditions. So the dissertation discusses the research of the concentration of sugar in blood of diabetes patients. In this part, this dissertation classified the patients into three levels. And it discusses the difference of the three levels. Then it classified the patients into other two levels. And it also shows the result of the difference of the two levels.