Nonlinear analysis of spatio-temporal receptive fields: III. RF-reconstruction from mean-field approximations

Thomas Wennekers

Contact the author: Please use for correspondence this email.
Submission date: 19. Mar. 2001
Pages: 8
published in: Neurocomputing, 44 (2002), p. 213-218 
DOI number (of the published article): 10.1016/S0925-2312(02)00433-2
Bibtex
Keywords and phrases: cortex, v1, orientation tuning, neural field equations, meanfield approximation, receptive field reconstruction

Abstract:
Nonlinear neural field models describe spatio-temporal cortical activation patterns by one or several layers of locally interconnected neurons. Impulse response functions in such models relate to dynamic receptive fields in the considered sensory domain. I recently presented a new method to calculate approximate amplitudes and tuning widths of such functions and demonstrated that the amplitude reponses can often already be obtained from a low-dimensional set of "meanfield-type" ordinary differential equations. In the present work I show how approximations of the full spatio-temporal response functions can be reconstructed from this meanfield approximation. The reconstruction method is applied to the dynamics of orientation tuning in V1 neurons (Ringach et al., Nature 387: 281-284, 1997) and it is argued that these data are consistent with feedback as well as feedforward models.