GENERAL NONINVASIVE BLOOD GLUCOSE PREDICTION METHOD BASED ON TIMING ANALYSIS

摘要

A general noninvasive blood glucose prediction method based on a timing analysis. The method comprises five steps of data inputting and preprocessing, feature value screening, a single feature value model based on a timing analysis, multi-feature value fusion, and noninvasive blood glucose prediction. New noninvasive test data is input during the noninvasive blood glucose prediction, and a blood glucose prediction sequence is calculated using related feature value information, the single-feature value model and a multi-feature value model obtained by means of modelling. The prediction method is easy to implement, and can overcome the delay between a change in a human physiological parameter and a change in blood glucose so as to obtain a more accurate noninvasive blood glucose test result. The prediction method is suitable for different noninvasive blood glucose test methods and has generality.