Title :
Classification of atrial enlargement using neural networks
Author :
Diery, A. ; Abbosh, Y. ; Thiel, D.V. ; Cutmore, T.R.H. ; Rowlands, D.
Author_Institution :
Griffith Sch. of Eng., Griffith Univ., Brisbane, QLD
Abstract :
The aim of this study was to classify using a neural network LAE into mildly, moderately, and severely abnormal from a subjectpsilas P-wave. Cardiological features, wavelet features, and a combination of both were used to train the neural networks. It was found features derived from the wavelet energy spectrum performed better than the cardiological features on the test cases.
Keywords :
cardiology; medical computing; neural nets; wavelet transforms; cardiological feature; left atrial enlargement classification; neural network; wavelet energy spectrum; wavelet feature; Cardiology; Electrocardiography; Feature extraction; Morphology; Neural networks; Pattern matching; Performance evaluation; Testing; Ultrasonic imaging; Wavelet analysis;
Conference_Titel :
Antennas, Propagation and EM Theory, 2008. ISAPE 2008. 8th International Symposium on
Conference_Location :
Kunming
Print_ISBN :
978-1-4244-2192-3
Electronic_ISBN :
978-1-4244-2193-0
DOI :
10.1109/ISAPE.2008.4735506
