PARALLEL PRINCIPAL COMPONENT NEURAL NETWORKS FOR CLASSIFICATION OF EVENT-RELATED POTENTIAL WAVE-FORMS

Authors
SVEINSSON JR BENEDIKTSSON JA STEFANSSON SB DAVIDSSON K
Citation
Jr. Sveinsson et al., PARALLEL PRINCIPAL COMPONENT NEURAL NETWORKS FOR CLASSIFICATION OF EVENT-RELATED POTENTIAL WAVE-FORMS, Medical engineering & physics, 19(1), 1997, pp. 15-20
Citations number
20
Categorie Soggetti
Engineering, Biomedical
Journal title
Medical engineering & physics → ACNP
ISSN journal
13504533
Volume
19
Issue
1
Year of publication
1997
Pages
15 - 20
Database
ISI
SICI code
1350-4533(1997)19:1<15:PPCNNF>2.0.ZU;2-A
Abstract
Artificial neural networks (ANNs) are discussed in terms of classifica tion of brain auditory event-related potentials (ERPs). A new ANN arch itecture for the classification of ERPs is proposed. The new architect ure is called the parallel principal component neural network (PPCNN). The use of the PPCNN for classification of ERP data obtained from bot h normal control subjects and chronic schizophrenic patients is discus sed. Experimental results are given. (C) 1997 Elsevier Science Ltd for IPEMB.