2-D spatial frequency dependence of VEPs: A neural network analysis

Author

Iezzi, Raymond, Jr. ; Micheli-Tzanakou, Evangelia

Author_Institution

Dept. of Biomed. Eng., Rutgers Univ., New Brunswick, NJ, USA

fYear

1993

fDate

18-19 Mar 1993

Firstpage

111

Lastpage

112

Abstract

An approach to black-box identification-function problems is introduced. Visual evoked potentials (VEPs) were recorded from normal subjects using stimulus patterns that evolved from a totally random distribution of intensities to an ordered distribution representing a bar. The contribution of all spatial frequencies to the amplitudes of the evoked potentials was studied by performing a 2-D FFT on the stimulus images. The FFT spectra were used as inputs to a neural network as a black box approach in order to study the VEP spatial frequency tuning curves of the subjects

Keywords

fast Fourier transforms; medical image processing; multilayer perceptrons; visual evoked potentials; 2-D FFT; 2-D spatial frequency dependence; ALOPEX; black-box identification-function problems; neural network analysis; stimulus patterns; visual evoked potentials; Biomedical engineering; Biomedical imaging; Educational institutions; Equations; Frequency conversion; Frequency dependence; Gratings; Image converters; Neural networks; Tuning;

fLanguage

English

Publisher

ieee

Conference_Titel

Bioengineering Conference, 1993., Proceedings of the 1993 IEEE Nineteenth Annual Northeast

Conference_Location

Newark, NJ

Print_ISBN

0-7803-0925-1

Type

conf

DOI

10.1109/NEBC.1993.404396

Filename

404396