Development And Application Of Evaluation Techniques In Optimum Treatment Of Hypertension

The control of hypertension is a global problem in public health. The goal for thetreatment of hypertension is to decrease the blood pressure, and the aim is to reduce thetotal risk of cardiovascular disease efficiently. At present, the control rate of hypertensionis still very low. One of the leading causes for this problem is the lack of effective methodto evaluate the effect of hypertension treatment.Applying the method of fuzzy multivariable comprehensive assessment, which is basedon the analysis of multi physiological parameters, the current study developed acomprehensive index (CI) to evaluate the state of cardiovascular system. The effect ofanti-hypertension treatment is evaluated by analyzing the change of cardiovascular stateof hypertensive patients. The model was improved by introducing the radial basisfunction neural network (RBF) to stratify the cardiovascular risk state into four levels.Accordingly, the following tasks were completed:1. The cardiovascular parameters were analyzed to screen the hypertension mostrelated parameters.2. Weighed value of each parameter in treatment evaluation was determined byanalytic hierarchy process and judgment matrix.3. Subjection function of each parameter was established.4. A model of fuzzy multivariable comprehensive assessment was developed andimproved by RBF neural network.1500 subjects were involved in the study and divided into two groups: 836 subjects werein hypertension group and 666 normotensive subjects were in control group. The resultsshow that averaged CI of hypertension subjects was 0.594±0.1329, which wasdecreased significantly as compared with those in the control group (0.712±0.0736,P＜0.1). The index CI was applied to stratify cardiovascular risk into no risk, low risk, midrisk and high risk. It identified the level of low and mid risk satisfactorily. Further, themodel was improved by applying RBF neural network and its total accuracy in riskstratification is 95.67%, i.e. the accuracy for classifying no risk, low risk, mid risk andhigh risk are 97.73%,96.59%,94% and 100%, respectively.The results of the study show that CI is available in the assessment of cardiovascularrisk state and effective in cardiovascular risk startification.