PARAFAC Analysis of Neural Correlates of Face Detection

Jhoanna Pérez-hidalgo-gato, V. Rodríguez, Eduardo Martínez-montes

Research output: Chapter in Book/Published conference outputOther chapter contribution

Abstract

Neural correlates of face processing have been largely studied, but more emphasis has been done in the identification of a particular face. Here we study the neural correlates of the N170 peak corresponding to the correct and incorrect detection of faces through the use of the Bayesian Model Averaging procedure. Moreover, different components of electrical sources are extracted with a PARAFAC analysis of the data. PARAFAC is a generalization of principal component analysis to deal with multidimensional data, offering as a great advantage unique decompositions. PARAFAC analysis of the three-dimensional data formed by the array of BMA inverse solutions for each subject and each experimental condition, provide of characteristic BMA sources with corresponding profiles for subjects and conditions. This allowed the identification of different and common sources for correct and incorrect detection of faces.
Original languageEnglish
Title of host publicationAdvances in Cognitive Neurodynamics ICCN 2007
EditorsRubin Wang, Enhua Shen, Fanji Gu
PublisherSpringer
Chapter77
Pages447-450
ISBN (Electronic)978-1-4020-8387-7
ISBN (Print)978-1-4020-8386-0
DOIs
Publication statusPublished - 2008

Fingerprint

Dive into the research topics of 'PARAFAC Analysis of Neural Correlates of Face Detection'. Together they form a unique fingerprint.

Cite this