Analysis And Application Of The Clinical Data Of New Typecerebral Edema Monitor System

Brain diseases is a great threat to the health of human life,in which cerebral edema is a serious one.If not treated in time,it may result in serious sequel,even endangering the life of the patient.Moreover,cerebral hemorrhage and cerebral infarction are the most severe ones among cerebral edema.But their surface symptoms are relatively similar,which usually leads to the difficulties of diagnosing.The current medical science mainly uses medical imaging technology which are represented by CT(Computed Tomography)and MRI(Magnetic Resonance Imaging)to carry out disease diagnosis.However,such a traditional method because of the equipment's requirements on space and time cannot deal with quick diagnosis in time.The fourth generation of noninvasive brain edema monitor produced by Chongqing Born Fuke Medical Equipment Cooperation Ltd is relatively portable and mobile,the accuracy of patient data it acquired has been verified in both the theory and practice.Based on the analysis of clinical data,this paper studies the establishment and the application of cerebral edema monitor model as well as the discrimination of several indicators on early cerebral hemorrhage and early cerebral infarction.In order to achieve the research objectives,this paper has done the following work:(1)Firstly,the original data of patients has been cleaned by the method of the cosine similarity and Euclidean distance.Researcher selected the sweep vectors which are universal representative for linear regression from the overall data to calculate a standard sweep vector under medical background.Experiments shows that the combination of the two methods performed very well in cleaning out the abnormal situation in patient data.(2)A number of new cerebral edema monitor models are suggested.First,a set of samples,each consisting of four components: the average of the raw monitoring data and the labeled volumes of cerebral edema,cerebral hematoma,and hydrocephalus of a patient in a monitor phase,are selected by means of the above mentioned data-cleaning methods.Second,a linear monitor model and a quadratic monitor model are established,respectively,by considering the average of the raw monitored data as the dependable variable.Third,a linear monitor model and a quadratic monitor model are introduced,respectively,by regarding the volume of cerebral edema as the dependable variable.Next,a set of indexes are proposed to assess the performance of a monitor model.Thereby,the second quadratic monitor model is found to possess an optimal performance index among the four models.Forth,a detailed introduction on the application of the second quadratic monitor model in the new type cerebral edema monitor system as well as the design and realization of the system,is proposed in this part.(3)Based on the observation of effective data,the analysis of medical knowledge and some previous studies,eight indicators for early cerebral hemorrhage and early cerebral infarction were proposed.And with the eight indicators,more than 2660 sets of historical data under four electrode mode were analyzed.Finally,an indicator with a higher degree of discrimination is presented.At the same time,the linear regression method that are repeated used in this paper is statistically tested,which proves that the conclusion has good statistical significance.