Bradycardia Prediction in Preterm Infants Using Nonparametric Kernel Density Estimation

Subhasish Das, Bahman Moraffah, Ayan Banerjee, Sandeep K. S. Gupta, Antonia Papandreou-Suppappola

Abstract: In this paper, we propose a statistical method to predict the onset of bradycardia in preterm infants without any prior knowledge. To model information on the QRS complex R wave, we exploit nonparametric methods to estimate the density. Our proposed method takes advantage of the kernel density estimator in order to provide a statistical guarantee of 95% accuracy. We also demonstrate our results through simulations to support our proposed method using preterm infant electrocardiogram (ECG) signals from a database. We show that the method achieves a 5% false alarm rate in predicting the onset of upcoming bradycardia events.