D. Gao, M. Madden, M. Schukat, D. Chambers, and G. Lyons (Ireland)
Neural Networks, Back Propagation, Bayesian, ECG.
This paper presents a diagnostic system for cardiac arrhyth mias from ECG data, using an Artificial Neural Network (ANN) classifier based on a Bayesian framework. The Bayesian ANN Classifier is built by the use of a logistic regression model and the back propagation algorithm. A dual threshold method is applied to determine the diagnosis strategy and suppress false alarm signals. The experimen tal results presented in this paper show that more than 90% prediction accuracy may be obtained using the improved methods in the study. It is hoped that the system can be further developed and fine-tuned for practical application.
Important Links:
Go Back
