How does the striate cortex begin the reconstruction of the visual world?

The "pincushion grid" illusion

The illusion generated by a "pincushion grid" is not predicted from the two-dimensional Fourier transform of the grid. This implies that the visual system may not perform two-dimensional Fourier transforms of observed patterns.

Spatial frequency-contingent color aftereffects

Two-dimensional Fourier analysis of checkerboards reveals that major components are at a 45 degree angle to the check edges. After adapting to chromatic checkerboards, subjects who viewed achromatic grating stimuli reported that complementary color aftereffects are aligned with spatial frequency components rather than with the edges in the pattern.

Periodic excitability changes across the receptive fields of complex cells in the striate and parastriate cortex of the cat

1. Complex cells in cortical areas 17 and 18 of the cat have been studied in response to narrow slits and edges moving across the receptive field in the preferred direction and also to stationary slits of different widths. 2. Average response histograms, recorded as a narrow slit was moved across the receptive field, displayed a periodic series of peaks above a base line level. The response histogram for most area 17 and 18 cells contained five principal peaks; sometimes one or two weaker peaks were present at receptive field borders. The histogram for one cell located at the area 17-18 border showed thirteen distinct peaks. Periodic response patterns were also generated as an extended edge was moved across the receptive field. Plots of cell responses versus slit width for stationary slits of different widths also indicated periodic response pattern. 3. The accuracy of determining the preferred slit orientation was the single most important requirement for demonstrating the periodic response pattern. Significant changes in the appearance of the periodic pattern occurred even upon 5 degrees rotations away from the preferred orientation. 4. Average response histograms were also studied over a wide range of moving slit velocities. The number of peaks across corresponding spacings within the recewptive field remained constant over a range of velocities. Response amplitudes, however, were velocity dependent. Thus the response peaks remain associated with fixed positions within visual space independent of stimulus velocity, even though temporal as well as spatial factors may be involved in response selectivity and the periodic modulation. The most striking periodic response histograms were generated at the velocities which produced the greatest cell firing rates. Area 17 complex cells responded well to velocities of less than 0-5 degrees to 6-0 degrees/sec, but cells in area 18 generally required higher velocities, sometimes as high as 20 degrees--30 degrees/sec, for a good response. 5. Spatial frequencies for the periodic component of the receptive field for area 17 cells in the central visual area covered a range of three octaves up to 5 cycles/degree, and area 18 cells included another octave on the low frequency side. The spatial frequency of a cell was found to be roughly inversely proportional to the receptive field width. Only a small sample of area 18 cells was studied, but these cells tended to represent low spatial frequencies and to respond selectively to high velocity stimuli...

Neural basis of orientation perception in primate vision

Orientational differences in human visual acuity can be related parametrically to the distribution of optimal orientations for the receptive fields of neurons in the striate cortex of the rhesus monkey. Both behavioral measures of acuity and the distribution of receptive fields exhibit maximums for stimuli horizontal or vertical relative to the retina; the effect diminishes with distance from the fovea. The anisotropy in the neuronal population and in visual acuity appear to be determined by postnatal visual experience.

Visual acuity and contrast sensitivity in patients with cerebral lesions

Spatial contrast sensitivity as a function of spatial frequency was measured in patients with cerebral lesions. In most of these patients visual acuity, as measured by the Snellen chart, was 20/30 or better, yet marked departures from normal contrast sensitivity were found. The greatest loss in contrast sensitivity occurred at high frequencies, but in one patient the loss was greatest in the midfrequency range. This finding lends support to the channel hypothesis of spatial contrast discrimination.

Gratings mask bars and bars mask gratings: visual frequency response to aperiodic stimuli

Gratings and bars produce unexpected mutual visual masking. A grating masks a bar much less than a bar masks a bar; and a bar masks a grating uniformly over the grating field. These effects suggest that neural populations selective for size and orientation may be involved in frequency analysis rather than in simple feature detection.