Discriminating brain activated area and predicting the stimuli performed using artificial neural network

Fuente: Redalyc
Salvato in:
Dettagli Bibliografici
Autore principale: Rafael do Espírito Santo
Natura: Artículo científico
Lingua:en
Pubblicazione: Universidade Nove de Julho 2007
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1876484997024579584
author Rafael do Espírito Santo
author_facet Rafael do Espírito Santo
contents Discriminating brain activated area and predicting the stimuli performed using artificial neural network Rafael do Espírito Santo Maria G. Moraes Martin João Ricardo Sato Ingeniería FMRI Paradigm Activation Classifier Neural networks In this work, a Multilayer Perceptron implementation – MLPusing functional Magnetic Resonance Imaging (fMRI) is usedto infer stimuli performed. Sets of images of brain activationwere generated by visual, auditory and finger tapping paradigmsin 54 healthy volunteers. These images were used fortraining the MLP network in a leave-one-out manner in orderto predict the paradigm that a subject performed by usingother images, so far unseen by the MLP network. The aim inthis paper is the exploring of the influence of the number of thePrincipal Component (PC) on the performance of the MLP inclassifying fMRI paradigms. The classifier´s performance wasevaluated in terms of the Sensitivity and Specificity, PredictionAccuracy and the area Az under the receiver operating characteristics(ROC) curve. From the ROC analysis, values of Az upto 1 were obtained with 60 PCs in discriminating the visualparadigm from the auditory paradigm. 2007 artículo científico 1678-5428 https://www.redalyc.org/articulo.oa?id=81050213 en http://www.redalyc.org/revista.oa?id=810 Exacta application/pdf Universidade Nove de Julho Exacta (Brasil) Num.2 Vol.5
format Artículo científico
id redalyc_81050213
institution Redalyc
language en
publishDate 2007
publisher Universidade Nove de Julho
spellingShingle Discriminating brain activated area and predicting the stimuli performed using artificial neural network
Rafael do Espírito Santo
Ingeniería
FMRI
Paradigm
Activation
Classifier
Neural networks
Discriminating brain activated area and predicting the stimuli performed using artificial neural network Rafael do Espírito Santo Maria G. Moraes Martin João Ricardo Sato Ingeniería FMRI Paradigm Activation Classifier Neural networks In this work, a Multilayer Perceptron implementation – MLPusing functional Magnetic Resonance Imaging (fMRI) is usedto infer stimuli performed. Sets of images of brain activationwere generated by visual, auditory and finger tapping paradigmsin 54 healthy volunteers. These images were used fortraining the MLP network in a leave-one-out manner in orderto predict the paradigm that a subject performed by usingother images, so far unseen by the MLP network. The aim inthis paper is the exploring of the influence of the number of thePrincipal Component (PC) on the performance of the MLP inclassifying fMRI paradigms. The classifier´s performance wasevaluated in terms of the Sensitivity and Specificity, PredictionAccuracy and the area Az under the receiver operating characteristics(ROC) curve. From the ROC analysis, values of Az upto 1 were obtained with 60 PCs in discriminating the visualparadigm from the auditory paradigm. 2007 artículo científico 1678-5428 https://www.redalyc.org/articulo.oa?id=81050213 en http://www.redalyc.org/revista.oa?id=810 Exacta application/pdf Universidade Nove de Julho Exacta (Brasil) Num.2 Vol.5
title Discriminating brain activated area and predicting the stimuli performed using artificial neural network
topic Ingeniería
FMRI
Paradigm
Activation
Classifier
Neural networks
url https://www.redalyc.org/articulo.oa?id=81050213