A neural network method for detecting unipolar activations during ventricular fibrillation

Author

Tohmaz, Abdul S. ; Blanchard, Susan M. ; Clark, Bryan P. ; Johnson, Eric E. ; Ideker, Raymond E.

Author_Institution

Dept. of Biol. & Agric. Eng., North Carolina State Univ., Raleigh, NC, USA

Volume

1

fYear

1995

fDate

20-25 Sep 1995

Firstpage

351

Abstract

An artificial neural network was trained to detect local activations during the first 2 sec of an induced ventricular fibrillation performed on five pigs. A rule based method (RBM) and methodology based on the transmembrane current (TCM) were used to validate the data for training/testing the network. With appropriate pre- and post-processing, the neural network yielded encouraging results when trained and tested on a portion of the data. A stop training technique was implemented to avoid over-training the network

Keywords

electrocardiography; medical signal processing; neural nets; 2 s; artificial neural network; local activations detection; neural network method; over-training avoidance; pigs; post-processing; pre-processing; rule based method; stop training technique; transmembrane current; unipolar activations detection; ventricular fibrillation; Agricultural engineering; Artificial neural networks; Electrodes; Fibrillation; Gold; Local activities; Monitoring; Neural networks; Testing; Training data;

fLanguage

English

Publisher

ieee

Conference_Titel

Engineering in Medicine and Biology Society, 1995., IEEE 17th Annual Conference

Conference_Location

Montreal, Que.

Print_ISBN

0-7803-2475-7

Type

conf

DOI

10.1109/IEMBS.1995.575145

Filename

575145