A NONLINEAR FUZZY LINGUISTIC PREDICTION MODEL FOR ACUTE HYPERGLYCEMIA USING CARDIAC ELECTROPHYSIOLOGICAL SIGNALS

Abstract

A nonlinear fuzzy linguistic prediction (NFLP) model for acute hyperglycemia prediction is proposed in this paper. The model used IF–THEN expressions which are human-readable and easy to understand. Using cardiac electrophysiological signals as the input, the model can predict actuation durations and concentrations of acute hyperglycemia. The prediction results are compared with the ones of four classical models which are partial least squares (PLS), least-square support vector machine (LSSVM), back-propagation neural network (BPNN) and Takagi–Sugeno (T–S) model. The results show that the proposed method has high prediction accuracy. The method can provide support for clinical diagnosis of acute hyperglycemia.