REM sleep eye movement counts correlate with visual imagery in dreaming: a pilot study.

Based on the findings of our previously published positron emission tomography study, we proposed that recorded eye movements during REM sleep are visually targeted saccades. In the present study, we examined the correlation between the number of eye movements in REM sleep (EM) and visual imagery in dreaming (V) and provided further support for our proposal. All the observations (N = 11) were made with one individual to eliminate interindividual variation and were made during the second REM sleep period to control for a time-of-night effect. V, with or without dream report length partialled out, was strongly associated with EM only in the 1-min interval immediately preceding awakening. The time course of the association suggests that the strong EM-V association reflects a phasic, localized activation of the eye-movement-control system in association with REM sleep eye movements.

Neuronal mechanisms of color categorization in areas V1, V2 and V4 of macaque monkey visual cortex.

A landmark study conducted by Berlin and Kay (Basic Color Terms, University of California Press, Berkeley, 1969, pp. 1-12) demonstrates that well-developed languages contain exactly 11 basic color terms. The basic colors (8 chromatic and 3 achromatic) are situated in specific locations of color space, suggesting a fixed relationship between specific hue and luminance. To determine the physiologic origins of the basic colors, we have studied the responses of cells in visual cortical areas V1, V2 and V4 of the behaving macaque monkey, using chromatic and achromatic stimuli of varying luminance. A total of 569 cells (291 from V1, 205 from V2, 73 from V4) were obtained, and classified as 'B' (bright; 43-50% of the total cells in each area), 'D' (dark; 6-12% of the total), and 'B/D' (bright/dark; 27-28% of the total) color or non-color cells according to each cell's color/luminance preference in relation to the neutral gray background. About two thirds of 'B' cells in each area were color specific, whereas the proportion of color cells in 'B/D' and 'D' categories was lower. In all three areas (v1, V2, V4), color cells with preferences for midspectral colors (such as yellow, lime and green) also preferred high luminance levels, while color cells with preferences for endspectral colors (such as red and blue) responded preferentially to luminance levels closer to background. The date provide evidence for categorical color perception within the visual system, as well as providing a physiological basis for the increased saliency of endspectral contours observed at equiluminance in psychophysical studies.

Color, orientation and cytochrome oxidase reactivity in areas V1, V2 and V4 of macaque monkey visual cortex.

Color and orientation processing in the macaque monkey first segregates into cytochrome oxidase (CO)-rich blobs and -poor interblobs of area V1, from where the two streams flow through areas V2 and V4. This parallel representation is believed to enhance processing speed by compartmentalizing tasks of similar kinds, though our knowledge of the mechanisms is still elementary. We have examined the interaction and separation of color and orientation processing in neurons (n = 569) of the macaque visual cortex (V1, V2, V4) on the basis of microelectrode recordings. In all three areas, neurons selective for midspectral (MS) colors (e.g., yellow, green) were also found to be more orientation selective than those preferring endspectral (ES) colors (e.g., blue, red). The majority of achromatic (AC) cells responsive to bright stimuli were also orientation selective. When locations of cells and penetration columns were correlated with cytochrome oxidase (CO) landmarks in V1 and V2, V1 interblob and V2 interstripe cells were found to be predominantly midspectral and oriented, while V1 blob and V2 thin stripe cells were found to be predominantly endspectral and non-oriented. Cells preferring dark colors were found to cluster in thick stripes in V2, and in columns in V4. Separate clustering of midspectral (MS) and endspectral (ES) systems in V4 was also noted. With the results shown in a companion paper (Behav. Brain Res., 76 (1996) 51-70), the present data indicate that the visual system appears to optimize color and spatial acuity by separating chromatic information into non-oriented endspectral and oriented midspectral components.

Sleep and dreams in Vietnam PTSD and depression.

The present study compares the sleep and dreams of three groups of subjects: 1) Vietnam veterans with posttraumatic stress disorder (PTSD) and major depression, 2) veterans with depression alone, and 3) veterans with neither PTSD nor depression (i.e., normal controls). Sleep recordings indicate only one significant difference between the PTSD/depressed and depressed alone groups: sleep latency was prolonged in the depressed alone patients compared with the other two groups. The two patient groups differed from controls in the manner already reported for depressed patients (decreased REM latency, increased REM density, reduced total sleep time, reduced sleep efficiency), with some of the differences significant only at the trend level. Dreams were obtained from REM awakenings. Dream recall rate and report length did not differ between groups. Mean anxiety level in dreams was less than 1 (mild) for all three groups, with major depression patients scoring significantly higher than controls. Dreams of PTSD/depressed patients were significantly less likely to be set in the present than dreams of the other two groups.

Localized and lateralized cerebral glucose metabolism associated with eye movements during REM sleep and wakefulness: a positron emission tomography (PET) study.

In order to study the neural substrate for eye movements during rapid eye movement (REM) sleep, we analyzed the positron emission tomography (18Fluorodeoxyglucose positron emission tomography) scan data obtained from normal subjects. Eye movement data were available on nine subjects studied during nighttime REM sleep and six control subjects studied during waking as they periodically moved their eyes. The number of eye movements during REM sleep was positively correlated with glucose metabolic rate in the areas corresponding to (a) the saccadic eye movement system (frontal eye field and dorsolateral prefrontal cortex, statistically significant only on the right side), (b) the midline attentional system (cingulate and medial frontal cortex, precuneus) and (c) the parietal visual spatial attentional system (bilateral superior parietal lobules, right inferior parietal lobule); and negatively correlated with relative metabolic rate in the left inferior parietal lobule. Positive correlations between waking eye movements and metabolic rate were observed in the same areas except inferior parietal lobule. Our results show that the same cortical areas are involved in eye movements in both REM sleep and wakefulness and suggest that REM sleep eye movements are saccadic scans of targets in the dream scene. Our data also suggest right hemispheric specialization in saccadic eye movement control and reciprocal inhibition in the contralateral homologous area during higher cortical functioning.

Local and global principles of striate cortical organization: an advanced model.

A model of the functional architecture in monkey striate cortex is proposed, based on recent data, in which the global structure is a result of the local structure. The local structure itself is organized around the cytochrome oxidase blobs. It is characterized by different anisotropic distributions of preferred orientations in upper and lower layers. In particular, the upper layers show a bias for "radial" orientations and the lower layers show a bias for "concentric" orientations (Bauer and Dow 1989). This organization may derive from the symmetry of the visual field and visual system which tends to be radial or concentric rather than cartesian. Functionally, the increased sensitivities of the different radial and concentric maps of orientation and movement detectors seem to be an adaptive fit to optic flow fields on the retina during complex movements of the subject in its environment.

Foveal tracking cells in the superior temporal sulcus of the macaque monkey.

Visual responses were recorded from neurons in the superior temporal sulcus (STS) of awake, behaving cynomolgus monkeys trained to fixate a small spot of light. Visual receptive fields, directionality, and responses during visual tracking were examined quantitatively for 50 cells in the foveal portion of the middle temporal (MT) visual area and surrounding cortex. Directionality indices and preferred directions for tracked and nontracked stimuli were compared. Eighteen cells (18/50 = 36%) were found to respond preferentially during tracking (tracking cells), 7 within MT, 9 in area FST on the floor of the STS, and 2 in unidentified areas. Three distinctly different tracking response profiles (VTS, VTO, and T) were observed. VTS and VTO cells had foveal receptive fields and gave directionally selective visual responses. VTS cells (3 in foveal MT, 6 in FST, 1 in an unidentified area) had a preferred visual direction that coincided with the preferred tracking direction, and began responding 50-100 ms before the onset of tracking. VTO cells (4 in foveal MT, 0 in FST, 1 in an unidentified area) had a preferred visual direction opposite to the preferred tracking direction, and began responding 0-100 ms after the onset of tracking. T cells (0 in MT, 3 in FST) had no visual responses and began responding simultaneously with the onset of tracking. It is suggested that this region of the brain could be the primary location for converting direction-specific visual responses into signals specifying at least the direction of an intended pursuit movement.

Representation of the fovea in the superior temporal sulcus of the macaque monkey.

The response properties of 633 neurons from striate and prestriate cortex were recorded in 3 hemispheres of two awake cynomolgus monkeys while they fixated or tracked a small spot of light. Of 254 penetrations located at 1 mm intervals, 39% were identifiable from visible electrolytic lesions or electrode tracks and were used to reconstruct the positions of all recording sites. A total of 226 cells were located in the superior temporal sulcus and 81 cells in area V1. The location and visuotopic organization of the foveal portion of the middle temporal (MT) visual area were determined in three hemispheres. MT was defined physiologically on the basis of direction-selectivity, receptive field size, and retinotopic organization. Of 170 MT neurons, most were motion sensitive, and 65% had a directionality index, (best-opposite)/best, of 0.6 or higher. MT was defined anatomically on the basis of myelin staining within the superior temporal sulcus (STS). On the posterior bank of the STS the physiologically defined border corresponded closely to a myelin border visible on our sections. Distinct myelin borders were not consistently identifiable on the anterior bank. The representation of the central fovea (eccentricities of 0-1 deg) was located partly on the floor, but mostly on the posterior bank of the STS at the extreme postero-lateral edge of MT. In all three hemispheres foveal MT extended onto the roof of a cleft formed between the posterior bank and a wide flattened area on the floor of the STS. This region lies 10-12 mm below the brain surface, measuring along a line normal to the surface at a point 2-3 mm antero-lateral to foveal V1. The area of MT was 6-9 mm2 for the central fovea (0-1 deg), 15-24 mm2 for the entire fovea (0-3 deg), and 28-40 mm2 including the fovea and parafovea (0-10 deg). A visuotopic map of central foveal V1 (0-1 deg) was obtained in one animal. The measured area of this representation was 116 mm2. Using published estimates of the total areas of cynomolgus MT and V1 (73 and 1200 mm2 respectively) the ratio of central foveal to total area was calculated to be 0.10 for both MT (7.5/73) and V1 (116/1200), indicating that the relative magnification of the foveal versus the peripheral visual field is preserved in the mapping of V1 onto MT. A separate representation of the central visual field was found immediately adjacent to foveal MT.(ABSTRACT TRUNCATED AT 400 WORDS)

Horizontal organization of orientation-sensitive cells in primate visual cortex.

In the visual cortex of the monkey the horizontal organization of the preferred orientations of orientation-selective cells follows two opposing rules: (1) neighbors tend to have similar orientation preferences, and (2) many different orientations are observed in a local region. We have described a classification for orientation maps based on the types of topological singularities and the spacing of these singularities relative to the cytochrome oxidase blobs. Using the orientation drift rate as a measure we have compared simulated orientation maps to published records of horizontal electrode recordings.

Complementary global maps for orientation coding in upper and lower layers of the monkey's foveal striate cortex.

A population of 269 cells recorded from the foveal representation of striate cortex in 2 rhesus macaque monkeys was examined for orientation preference as a function of receptive field position relative to the center of gaze. Cells recorded in supragranular and infragranular layers were segregated and compared. Within the foveolar region (0.0-0.5 degrees) supragranular cells showed a vertical bias which was not evident in the infragranular layers. At larger eccentricities (0.5-2.5 degrees) supragranular cells showed a radial bias (preferred orientation points toward the center of gaze), whereas infragranular cells showed a concentric bias (preferred orientation is tangent to a circle around the center of gaze). These results are consistent with our earlier report of an orientation shift between the supragranular and infragranular layers (Bauer et al. 1980, 1983; Dow and Bauer 1984). The diagonal orientation bias which we noted earlier (Bauer et al. 1980; Dow and Bauer 1984) in supragranular cells at eccentricities between 0.5 and 2.5 degrees can be explained by the radial bias, combined with a tendency for recording sites to favor receptive field locations closer to the diagonal meridia than to either the horizontal or vertical meridia. Given other evidence that upper layer cells in macaque striate cortex tend to show either orientation or color selectivity, while lower layer cells tend to show movement sensitivity (Dow 1974; Livingstone and Hubel 1984), the present data suggest a functional dichotomy between a supragranular system involved in fixational eye movements and pattern vision and an infragranular system activated primarily by optical flow fields during ambulation.

Horizontal segregation of color information in the middle layers of foveal striate cortex.

The present study examines the chromatic organization of foveal striate cortex in the awake monkey by means of a series of microelectrode penetrations made as perpendicular as possible to the layers. The study includes 79 penetrations and 261 cells, of which 218 were tested systematically for color selectivity. Detailed analyses are conducted on a subset of 41 penetrations, which included 164 color-tested cells, an average of four cells per penetration. The penetrations were divided into two major categories on the basis of orientation selectivity testing. One group of penetrations contained at least one nonoriented cell in the first 600 microns of microelectrode trajectory (upper layers), whereas the other group of penetrations contained only oriented cells in the first 600 microns of microelectrode trajectory. The two groups were called N (nonoriented) and O (oriented), respectively. Analysis of the color properties of cells in N and O penetrations revealed that middle layer cells in N penetrations showed poor responses to white light, and color preferences for endspectral wavelengths, i.e., red or blue. Middle layer cells in O penetrations, by contrast, responded well to white light and to midspectral wavelengths. There were two subgroups of N penetrations, characterized by predominantly red (NR) or blue (NB) sensitivity in the middle layers. O penetrations could likewise be divided up into three subgroups (OG, OY, OW), characterized, respectively, by predominant sensitivity to greenish wavelengths (490-540 nm), yellowish wavelengths (550-600 nm), or white (i.e., all colors). Despite the identification of five subgroups, similarities between NR and OY, between NB and OG, and between OY and OW subgroups might be consistent with a continuum. The middle layers of N penetrations contained a unique class of cells with excitatory responses restricted to the two spectral ends, endspectral double-peak cells. A model is proposed for the color organization of layer 4 in foveal striate cortex, with red and blue zones in register with alternate cytochrome oxidase "blobs" of layers 2 and 3, white zones in register with interblob centers, and yellow and green zones in between.

Color cell groups in foveal striate cortex of the behaving macaque.

Color-tuning curves were obtained for 218 cells in the foveal striate cortex of behaving macaques. Each cell was tested with its optimal spatial stimulus. Test colors (14 interference filters, 4 Wratten filters, and white) were matched for human photopic luminosity and presented at luminance levels sufficient to induce vigorous responding from most cells. One hundred eighty-four cells were selected for further analysis on the basis of a color-tuning index. Of these, 130 with tuning curves that correlated well (0.9 or better) with other tuning curves were studied in detail. Individual cells were found with peak responses to every color tested. Sixty-three tuning curves fell into the six largest cross-correlation groups, containing 15, 14, 12, 9, 7, and 6 cells, with mean tuning-curve peaks at 450, 656, 656, 506, 577, and 506 nm, respectively. Cross-correlation groups having the same peak location (656 nm, 506 nm) were distinguishable on the basis of tuning-curve width. Response patterns, cone input estimates, and comparison with human psychophysics suggest that two of these cell groups function as an opponent pair processing the colors red and green. Two other cell groups process the colors blue and yellow but show less well-developed opponency. Microdrive depth readings, correlated with histological lesion sites, show these "red," "green," "blue," and "yellow" cells to be most common in layer 4 of the striate cortex.

The mapping of visual space onto foveal striate cortex in the macaque monkey.

Visuotopic maps of foveal striate cortex have been obtained by means of single cell recordings from four hemispheres in two awake, behaving macaque monkeys. The numbers of successful separate striate penetration sites in the four hemispheres were 42, 58, 81, and 61, for a total of 242. The resolution of the maps is 10 min of visual angle, nearly an order of magnitude finer than previous maps. No striate receptive field center was found more than 5 min into the ipsilateral visual field. The four maps were sufficiently compatible with one another that they could be combined into one. There are only minor magnification differences between the right and left hemispheres and between the upper and lower quadrants. There is a vertical/horizontal magnification anisotropy of about 1.5:1 in central foveal cortex (0 to 20 min). The composite map can be approximated by the complex logarithmic equation, w = 7.7 * ln (x + iy + 0.33), where w is expressed in millimeters and x and y are expressed in degrees.

Retinotopy and orientation columns in the monkey: a new model.

A model is presented in which orientation columns arise directly out of retinotopy. According to the model, iso-orientation lines are arrayed radially around nodal centers which correspond to cytochrome oxidase patches. The nodal centers form a square matrix superimposed upon the map of ocular dominance stripes. In the supragranular layers horizontal iso-orientation lines run down the centers of ocular dominance stripes, with vertical iso-orientation lines crossing perpendicularly. Diagonal orientations (45 degrees and 135 degrees) are represented as alternating iso-orientation zones at the centers of the interstices in the matrix (internodal centers). Preferred orientations in the infragranular layers are reversed with respect to the supragranular layers. The model is consistent with new data concerning ocularity and preferred orientation in systematic penetrations through striate cortex, and helps to explain some previously puzzling features of the relationship between ocular dominance columns, orientation columns and retinotopy.

Orientation shift between upper and lower layers in monkey visual cortex.

Microelectrode penetrations nearly normal to the layers of foveal striate cortex in awake, behaving monkeys reveal a shift in orientation preference between cells in the upper and lower layers. Mean shift size for 57 penetrations is 54.8 degrees, with 70% of the penetrations showing shifts of 45-90 degrees. Marking lesions localize the shift to the border between layers 4C and 5. The data are suggestive of inhibition between the upper and lower layers within an "orientation column".

Magnification factor and receptive field size in foveal striate cortex of the monkey.

Receptive field size and magnification have been studied in striate cortex of awake, behaving rhesus monkeys at visual eccentricities in the range of 5-160 min. The major findings that emerge are (1) magnification in the foveola achieves values in the range of 30 mm/deg, (2) mean field size is not proportional to inverse magnification in contrast with previous reports, and (3) the product, magnification X aggregate field size, is greater in central vision than in peripheral vision. Thus, a point of light projected onto foveal retina is "seen" by larger numbers of striate cortical cells than a point of light projected onto peripheral retina. Implications of these findings for visual localization and two-point discrimination are discussed.

Behavioral enhancement of visual responses of prestriate neurons of the rhesus monkey.

Neurons in the superior colliculus, striate cortex, frontal eye fields, and posterior parietal cortex of the monkey respond to visual stimuli. Many of these cells discharge more intensely to a stimulus when it is to be the target for a saccadic eye movement than when fixation is maintained. We have demonstrated that such enhancement of the visual response is also present for cells in prestriate cortex. The prestriate effect is a modulation of the visual response and not a concomitant of oculomotor activity. It is present for eye movements away from as well as into the visual receptive field and is thus similar to that seen in striate cortex and different from that studied in the superior colliculus, frontal eye fields, and posterior parietal cortex. The visual responses of many prestriate cells habituate with repeated stimulation. When the monkey makes saccadic eye movements to a stimulus that is eliciting only a habituated response, the enhancement acts as a dishabituation which persists throughout the eye movement trials.