Arrhythmia Identification from ECG Signals with a Neural Network Classifier Based on a Bayesian Framework
Lyons, Gerard J.
Madden, Michael G.
MetadataShow full item record
This item's downloads: 1 (view details)
"Arrhythmia Identification from ECG Signals with a Neural Network Classifier Based on a Bayesian Framework" , Dayong Gao, Michael G. Madden, Michael Schukat, Des Chambers, and Gerard Lyons. Work-in-Progress track of AI-2004, the Twenty-fourth SGAI International Conference on Innovative Techniques and Applications of Artificial Intelligence, December 2004.
This paper presents an ANN-based diagnostic system for arrhythmia using Neural Network Classifier with Bayesian framework by time series biosignals. The Neural Network Classifier is built by the use of logistic regression model and back propagation algorithm. The prediction per-formance in training and test phases is evaluated by the False Rate. The dual threshold method is applied to determine diagnosis strategy and suppress false alarm signals. The results show that more than 90% prediction accuracy could be obtained using the improved methods in the study. Hopefully, the system can be further developed and fine-tuned for practical application.