Stimulus Dependence of Gamma Oscillations in Human Visual Cortex.

A striking feature of some field potential recordings in visual cortex is a rhythmic oscillation within the gamma band (30-80 Hz). These oscillations have been proposed to underlie computations in perception, attention, and information transmission. Recent studies of cortical field potentials, including human electrocorticography (ECoG), have emphasized another signal within the gamma band, a nonoscillatory, broadband signal, spanning 80-200 Hz. It remains unclear under what conditions gamma oscillations are elicited in visual cortex, whether they are necessary and ubiquitous in visual encoding, and what relationship they have to nonoscillatory, broadband field potentials. We demonstrate that ECoG responses in human visual cortex (V1/V2/V3) can include robust narrowband gamma oscillations, and that these oscillations are reliably elicited by some spatial contrast patterns (luminance gratings) but not by others (noise patterns and many natural images). The gamma oscillations can be conspicuous and robust, but because they are absent for many stimuli, which observers can see and recognize, the oscillations are not necessary for seeing. In contrast, all visual stimuli induced broadband spectral changes in ECoG responses. Asynchronous neural signals in visual cortex, reflected in the broadband ECoG response, can support transmission of information for perception and recognition in the absence of pronounced gamma oscillations.

Abnormal retinotopic representations in human visual cortex revealed by fMRI.

The representation of the visual field in early visual areas is retinotopic. The point-to-point relationship on the retina is therefore maintained on the convoluted cortical surface. Functional magnetic resonance imaging (fMRI) has been able to demonstrate the retinotopic representation of the visual field in occipital cortex of normal subjects. Furthermore, visual areas that are retinotopic can be identified on computationally flattened cortical maps on the basis of positions of the vertical and horizontal meridians. Here, we investigate abnormal retinotopic representations in human visual cortex with fMRI. We present three case studies in which patients with visual disorders are investigated. We have tested a subject who only possesses operating rod photoreceptors. We find in this case that the cortex undergoes a remapping whereby regions that would normally represent central field locations now map more peripheral positions in the visual field: In a human albino we also find abnormal visual cortical activity. Monocular stimulation of each hemifield resulted in activations in the hemisphere contralateral to the stimulated eye. This is consistent with abnormal decussation at the optic chiasm in albinism. Finally, we report a case where a lesion to white matter has resulted in a lack of measurable activity in occipital cortex. The activity was absent for a small region of the visual field, which was found to correspond to the subject's field defect. The cases selected have been chosen to demonstrate the power of fMRI in identifying abnormalities in the cortical representations of the visual field in patients with visual dysfunction. Furthermore, the experiments are able to show how the cortex is capable of modifying the visual field representation in response to abnormal input.

Visual areas and spatial summation in human visual cortex.

Functional MRI measurements can securely partition the human posterior occipital lobe into retinotopically organized visual areas (V1, V2 and V3) with experiments that last only 30 min. Methods for identifying functional areas in the dorsal and ventral aspect of the human occipital cortex, however, have not achieved this level of precision; in fact, different laboratories have produced inconsistent reports concerning the visual areas in dorsal and ventral occipital lobe. We report four findings concerning the visual representation in dorsal regions of occipital cortex. First, cortex near area V3A contains a central field representation that is distinct from the foveal representation at the confluence of areas V1, V2 and V3. Second, adjacent to V3A there is a second visual area, V3B, which represents both the upper and lower quadrants. The central representation in V3B appears to merge with that of V3A, much as the central representations of V1/2/3 come together on the lateral margin of the posterior pole. Third, there is yet another dorsal representation of the central visual field. This representation falls in area V7, which includes a representation of both the upper and lower quadrants of the visual field. Fourth, based on visual field and spatial summation measurements, it appears that the receptive field properties of neurons in area V7 differ from those in areas V3A and V3B.

Scene illuminant classification: brighter is better.

Knowledge of the scene illuminant spectral power distribution is useful for many imaging applications, such as color image reproduction and automatic algorithms for image database applications. In many applications accurate spectral characterization of the illuminant is impossible because the input device acquires only three spectral samples. In such applications it is sensible to set a more limited objective of classifying the illuminant as belonging to one of several likely types. We describe a data set of natural images with measured illuminants for testing illuminant classification algorithms. One simple type of algorithm is described and evaluated by using the new data set. The empirical measurements show that illuminant information is more reliable in bright regions than in dark regions. Theoretical predictions of the algorithm's classification performance with respect to scene illuminant blackbody color temperature are tested and confirmed by using the natural-image data set.

Visualization and measurement of the cortical surface.

Much of the human cortical surface is obscured from view by the complex pattern of folds, making the spatial relationship between different surface locations hard to interpret. Methods for viewing large portions of the brain's surface in a single flattened representation are described. The flattened representation preserves several key spatial relationships between regions on the cortical surface. The principles used in the implementations and evaluations of these implementations using artificial test surfaces are provided. Results of applying the methods to structural magnetic resonance measurements of the human brain are also shown. The implementation details are available in the source code, which is freely available on the Internet.

Computational neuroimaging of human visual cortex.

Functional magnetic resonance imaging is a new neuroimaging method for probing the intact, alert, human brain. With this tool, brain activity that has been hidden can now be measured. Recent advances in measuring and understanding human neural responses underlying motion, color, and pattern perception are reviewed. In individual human brain, we can now identify the positions of several retinotopically organized visual areas; measure retinotopic organization within these areas; identify the location of a motion-sensitive region in individual brains; measure responses associated with contrast, color, and motion; and measure effects of attentional modulation on visually evoked responses. By framing experiments and analyses as questions about visual computation, these neuroimaging measurements can be coupled closely with those from other basic vision-science methods.

Topographic organization of human visual areas in the absence of input from primary cortex.

Recently, there has been evidence for considerable plasticity in primary sensory areas of adult cortex. In this study, we asked to what extent topographical maps in human extrastriate areas reorganize after damage to a portion of primary visual (striate) cortex, V1. Functional magnetic resonance imaging signals were measured in a subject (G.Y.) with a large calcarine lesion that includes most of primary visual cortex but spares the foveal representation. When foveal stimulation was present, intact cortex in the lesioned occipital lobe exhibited conventional retinotopic organization. Several visual areas could be identified (V1, V2, V3, V3 accessory, and V4 ventral). However, when stimuli were restricted to the blind portion of the visual field, responses were found primarily in dorsal extrastriate areas. Furthermore, cortex that had formerly shown normal topography now represented only the visual field around the lower vertical meridian. Several possible sources for this reorganized activity are considered, including transcallosal connections, direct subcortical projections to extrastriate cortex, and residual inputs from V1 near the margin of the lesion. A scheme is described to explain how the reorganized signals could occur based on changes in the local neural connections.

Perceived speed of colored stimuli.

The influence of contrast and color on perceived motion was measured using a speed-matching task. Observers adjusted the speed of an L cone contrast pattern to match that of a variety of colored test patterns. The dependence of speed on test contrast was the same for all test colors measured, differing only by a sensitivity factor. This result suggests that the reduced apparent speed of low contrast targets and certain colored targets is caused by a common cortical mechanism. The cone contrast levels that equate perceived speed differ substantially from those that equate visibility. This result suggests that the neural mechanisms governing speed perception and visibility differ. Perceived speed differences caused by variations in color can be explained by color responses that are characteristic of motion-selective cortex.

Color signals in human motion-selective cortex.

The neural basis for the effects of color and contrast on perceived speed was examined using functional magnetic resonance imaging (fMRI). Responses to S cone (blue-yellow) and L + M cone (luminance) patterns were measured in area V1 and in the motion area MT+. The MT+ responses were quantitatively similar to perceptual speed judgments of color patterns but not to color detection measures. We also measured cortical motion responses in individuals lacking L and M cone function (S cone monochromats). The S cone monochromats have clear motion-responsive regions in the conventional MT+ position, and their contrast-response functions there have twice the responsivity of S cone contrast-response functions in normal controls. But, their responsivity is far lower than the normals' responsivity to luminance contrast. Thus, the powerful magnocellular input to MT+ is either weak or silent during photopic vision in S cone monochromats.

Color signals in area MT of the macaque monkey.

The relationship between the neural processing of color and motion information has been a contentious issue in visual neuroscience. We examined this relationship directly by measuring neural responses to isoluminant S cone signals in extrastriate area MT of the macaque monkey. S cone stimuli produced robust, direction-selective responses at most recording sites, indicating that color signals are present in MT. While these responses were unequivocal, S cone contrast sensitivity was, on average, 1.0-1.3 log units lower than luminance contrast sensitivity. The presence of S cone responses and the relative sensitivity of MT neurons to S cone and luminance signals agree with functional magnetic resonance imaging (fMRI) measurements in human MT+. The results are consistent with the hypothesis that color signals in MT influence behavior in speed judgment tasks.

Trichromatic opponent color classification.

Stimuli varying in intensity and chromaticity, presented on numerous backgrounds, were classified into red/green, blue/yellow and white/black opponent color categories. These measurements revealed the shapes of the boundaries that separate opponent colors in three-dimensional color space. Opponent color classification boundaries were generally not planar, but their shapes could be summarized by a piecewise linear model in which increment and decrement color signals are combined with different weights at two stages to produce opponent color sensations. The effect of background light on classification was largely explained by separate gain changes in increment and decrement cone signals.

Retinotopic organization in human visual cortex and the spatial precision of functional MRI.

A method of using functional magnetic resonance imaging (fMRI) to measure retinotopic organization within human cortex is described. The method is based on a visual stimulus that creates a traveling wave of neural activity within retinotopically organized visual areas. We measured the fMRI signal caused by this stimulus in visual cortex and represented the results on images of the flattened cortical sheet. We used the method to locate visual areas and to evaluate the spatial precision of fMRI. Specifically, we: (i) identified the borders between several retinotopically organized visual areas in the posterior occipital lobe; (ii) measured the function relating cortical position to visual field eccentricity within area V1; (iii) localized activity to within 1.1 mm of visual cortex; and (iv) estimated the spatial resolution of the fMRI signal and found that signal amplitude falls to 60% at a spatial frequency of 1 cycle per 9 mm of visual cortex. This spatial resolution is consistent with a linespread whose full width at half maximum spreads across 3.5 mm of visual cortex.

Creating connected representations of cortical gray matter for functional MRI visualization.

We describe a system that is being used to segment gray matter from magnetic resonance imaging (MRI) and to create connected cortical representations for functional MRI visualization (fMRI). The method exploits knowledge of the anatomy of the cortex and incorporates structural constraints into the segmentation. First, the white matter and cerebral spinal fluid (CSF) regions in the MR volume are segmented using a novel techniques of posterior anisotropic diffusion. Then, the user selects the cortical white matter component of interest, and its structure is verified by checking for cavities and handles. After this, a connected representation of the gray matter is created by a constrained growing-out from the white matter boundary. Because the connectivity is computed, the segmentation can be used as input to several methods of visualizing the spatial pattern of cortical activity within gray matter. In our case, the connected representation of gray matter is used to create a flattened representation of the cortex. Then, fMRI measurements are overlaid on the flattened representation, yielding a representation of the volumetric data within a single image. The software is freely available to the research community.

Color appearance of mixture gratings.

We have examined how color appearance varies with spatial pattern. Subjects set color-matches between a uniform, 2 deg matching field and bars within squarewave patterns (1,2 and 4 c/deg) or the superposition of these squarewaves. The matches were set using squarewaves and squarewave mixtures with many different colors and contrasts. The color-matches satisfied the basic properties of a linear system to within a tolerance of twice the precision of repeated matches. Matches satisfied contrast-homogeneity: the contrast of the matching field was proportional to the contrast of the squarewave pattern or the mixture of squarewave patterns. Matches also satisfied pattern-superposition: if a bar in one squarewave matched one uniform field, and a bar in a second squarewave matched a second uniform field, the superposition of the two squarewave bars matched the superposition of the uniform matching fields. Matches are predicted by a model in which the color at a location is predicted by the responses of three linear, pattern-color separable mechanisms. As the individual mechanisms are pattern-color separable, meaningful pattern and color-responsivity functions can be estimated for each of the mechanisms. The estimated color-responsivity functions, based only on asymmetric color-matches, have an opponent-colors organization.

Seeing gray through the ON and OFF pathways.

Color appearance judgments revealed fundamental differences in visual processing of incremental and decremental lights. First, the balance of cone activation required for a light to appear achromatic was different for increments and decrements (Judd, 1940; Helson & Michels, 1948). Second, adaptation--the visual system's adjustment to background light--affected achromatic decrements more than increments. Third, the regulation of adaptation for incremental and decremental stimuli depended differently on background signals from the three cone types. We interpret these asymmetries as differences in mechanisms of adaptation in the ON and OFF pathways, and suggest that they evolved to accommodate the range and physical sources of color signals in the two pathways.

Pattern-color separable pathways predict sensitivity to simple colored patterns.

We have studied how contrast threshold sensitivity depends jointly on pattern and color. We measured sensitivity to colored Gabor patches from 0.5 to 8 c/deg. At each spatial frequency, we measured in many different color directions. We analyze the sensitivity measurements using a series of nested models. We conclude that a model consisting of three pattern-color separable mechanisms predicts detection performance nearly as well as fitting psychometric functions independently. We derive the pattern and color sensitivities of the separable mechanisms from the experimental data. Two derived mechanisms are spatially lowpass and spectrally color-opponent. The third mechanism is spatially bandpass and spectrally broadband.

Photoreceptor sensitivity changes explain color appearance shifts induced by large uniform backgrounds in dichoptic matching.

Photoreceptor sensitivity changes explained the effect of large uniform backgrounds on the color appearance of small targets in a dichoptic asymmetric color matching experiment. Subjects viewed in each eye a target superimposed on a large background. The backgrounds presented to the two eyes had different spectral compositions. Subjects adjusted the target seen by the right eye to match the appearance of the target seen by the left eye. Receptor sensitivity changes explained the effect of numerous adapting backgrounds on the color appearance of many targets with high precision. Post-receptoral sensitivity changes provided a poorer account of the data. The apparent sensitivity of each receptor class varied inversely with changes in background light absorbed by that receptor class, but did not depend on background light absorbed by the other two receptor classes.

Appearance of colored patterns: pattern-color separability.

We have measured how color appearance of square-wave bars varies with stimulus strength and spatial frequency. Observers adjusted the color of a uniform patch to match the color appearance of the bars in square-wave patterns. We used low-to-moderate square-wave patterns, from 1 to 8 cycles per degree (c/deg). The matches are not photoreceptor matches but rather are established at more central neural sites. The signals at the putative central sites obey several simple regularities. The cone contrast of the uniform patch is proportional to square-wave stimulus strength (color homogeneity) and additive with respect to the superposition of equal-frequency square waves containing different colors (color superposition). We use the asymmetric matches to derive, from first principles, three pattern-color-separable appearance pathways. The matches are explained by two spectrally opponent, spatially low-pass mechanisms and one spectrally positive, spatially bandpass mechanism. The spectral mechanisms that we derive are similar to luminance and opponent mechanisms that are derived with entirely different experimental methods.

Color appearance: the effects of illumination and spatial pattern.

The color we perceive at each point in an image depends on information spread across the three spatial arrays of cone photoreceptors. I describe experiments aimed at clarifying how information is integrated across the spatial arrays to yield a color experience. We have found that changes of color appearance due to changes of the ambient illumination and the pattern's spatial frequency can be described by using a simple set of optical and neural transformations. Each transformation can be thought of as having two parts. First, the transformation converts the color representation into a new coordinate frame that is independent of the image contents. Second, the transformation scales the neural responses in the new coordinate frame by a gain factor that depends on the image contents.