DocumentCode :
2339648
Title :
Eigenvector methods for automated detection of time-varying biomedical signals
Author :
Guler, Inan ; Ubeyli, Elif Derya
Author_Institution :
Dept. of Electron. & Comput. Educ., Gazi Univ.
fYear :
0
fDate :
0-0 0
Abstract :
In this paper, we present the automated diagnostic systems for time-varying biomedical signals classification and determine their accuracies. The combined neural network (CNN) and mixture of experts (ME) were tested and benchmarked for their performance on the classification of the studied time-varying biomedical signals (ophthalmic arterial Doppler signals and electroencephalogram signals). Decision making was performed in two stages: feature extraction by eigenvector methods and classification using the classifiers trained on the extracted features. The purpose was to determine an optimum classification scheme for the problem and also to infer clues about the extracted features. Our research demonstrated that the power levels of power spectral density (PSD) estimations obtained by the eigenvector methods are the valuable features which are representing the time-varying biomedical signals and the CNN and ME trained on these features achieved high classification accuracies
Keywords :
Doppler measurement; eigenvalues and eigenfunctions; electroencephalography; eye; feature extraction; medical diagnostic computing; medical signal processing; neural nets; signal classification; automated diagnostic systems; combined neural network; eigenvector methods; electroencephalogram signals; feature extraction; mixture of experts; ophthalmic arterial Doppler signals; power spectral density estimations; time-varying biomedical signals classification; Biomedical engineering; Cellular neural networks; Computer science education; Data mining; Epilepsy; Feature extraction; Frequency estimation; Neural networks; Signal processing; Systems engineering education; Combined neural network (CNN); Eigenvector methods; Mixture of experts (ME); Time-varying biomedical signals;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computational Intelligence Methods and Applications, 2005 ICSC Congress on
Conference_Location :
Istanbul
Print_ISBN :
1-4244-0020-1
Type :
conf
DOI :
10.1109/CIMA.2005.1662296
Filename :
1662296
Link To Document :
بازگشت