Radial Basis Function Neural Network for Classification of Quantitative EEG in Patients with Advanced Chronic Renal Failure

摘要

In this study we investigate the potential of applying the radial basis function (RBF) neural network architecture for the classification of patients with chronic renal failure (CRF) through quantitative parameters derived from EEG. To provide an objective EEG assessment of cerebral disturbances in CRF, we set up and tested a procedure of classification based on artificial neural networks (ANN) using RBF trained with quantitative parameters derived from EEG. A set sample was prepared based on EEG of 17 patients and 18 age-matched control subjects. Quantitative EEG (qEEG) found significant differences between groups. Accuracy of ANN-based classification in this set was 86.6%. Our results indicate that a classification system based on RBF neural networks may help in the automation of EEG analysis for diagnosis and prospective clinical evaluation of CRF patients.