Detection of nocturnal hypoglycemic episodes (natural occurrence) in children with Type 1 diabetes using an optimal Bayesian neural network algorithm

Abstract

Hypoglycemia or low blood glucose is dangerous and can result in unconsciousness, seizures and even death. It is a common and serious side effect of insulin therapy in patients with diabetes. HypoMon is a non-invasive monitor that measures some physiological parameters continuously to provide detection of hypoglycemic episodes in Type 1 diabetes mellitus patients (T1DM). Based on heart rate and corrected QT interval of the ECG signal, we have continued to develop Bayesian neural network detection algorithms to recognize the presence of hypoglycemic episodes. From a clinical study of 16 children with T1DM, natural occurrence of nocturnal hypoglycemic episodes are associated with increased heart rates (1.033+/-0.242 vs. 1.082+/-0.298, P0.06) and increased corrected QT intervals (1.031+/-0.086 vs. 1.060+/-0.084, P0.001). The overall data were organized into a training set (8 patients) and a test set (another 8 patients) randomly selected. Using the optimal Bayesian neural network with 10 hidden nodes which was derived from the training set with the highest log evidence, the sensitivity (true positive) value for detection of hypoglycemia in the test set is 89.2%.