Right Visual Field
Figure 7 Visual field mapping in human occipital cortex. (A-C, A'-C'): The cortical representation of polar angle in the visual field in three subjects. The stimulus depicted at left was a counterphase flickering (8 Hz), checkered hemifield rotated slowly about the fixation point. For display purposes, the color code shows brain sites that were maximally activated when the center of the hemifield was within 45° of the horizontal meridian (white), the superior vertical meridian (dark gray), or the inferior vertical meridian (light gray). (A-C) 3-dimensional surface models of the occipital lobe plus adjacent parietal cortex, medial view, anterior to left. (A'-C'): Cortical flat maps showing the same data. White dashed lines at the edges of flat maps show margins of slit running from the occipital pole along the lateral occipital sulcus. Upper and lower halves of the dashed lines would be juxtaposed along the hidden lateral aspect of the 3-dimensional models shown in A-C. (D-F, D'-F'): Cortical representation of visual field eccentricity for the same three subjects. Stimulus: Flickering, checkered annulus that expanded slowly from 1.4° to 24°. Color code (gray to white) represents activation by six groups of three successive annuli. Inclusive eccentricities (degrees): 1.4-4.5, 2.3-7.6, 3.7-12.4, 6.3-20.8, 10.5-30, and 17.1-30. The coloring of the hemifield in the legend is only approximate; true mean eccentricities are shown at bottom. Right: Composite retinotopic grid (middle) combining data across subjects for vertical and horizontal meridian representations (right top) plus eccentricity (right bottom) displayed on a flat map of subject A. Lines in meridia and eccentricity maps pass through centers of corresponding meridian and isoeccentricity domains in A'-F'. Note multiple representations of meridia indicated at right of the composite and meridia maps. Abbreviations: HM, horizontal meridian representations; VM, vertical meridian representations (orange, lower field; yellow, upper field); Cal S, calcarine sulcus; Coll S, collateral sulcus; MT+, human middle temporal visual area and neighbors; POS, parieto-occipital sulcus; PVA, parietal visual areas. (Adapted from DeYoe etal., 1996, Proc. Natl. Acad. Sci. U.S.A. 93,2382-2386.)
has also been shown to contain both upper and lower field representations, though in other respects this area may be functionally different from V3A in the macaque. Therefore, the designation V3A should be considered provisional for humans.
The anterolateral border of area VP (V3v) is formed by a second representation of the superior vertical meridian (VMs). In the macaque, this forms the border between area VP and the ventral half of V4. Core-spondingly, the human area anterolateral to VP has also been designated as V4v. In both humans and macaques, V4v contains a representation of the superior field quadrant. In macaques, however, this is paired with an inferior quadrant representation located dorsally in the prelunate gyrus (cf Figs. 6F and 6G). This dorsal division, V4d, borders area V3A along an inferior vertical meridian representation. Human imaging studies have not identified a consistent focus dorsally that would correspond topographically and functionally to macaque V4d.
Anterior and lateral to V4v, near the occipitotem-poral junction, lies an area, like V3A, that contains a complete representation of both the upper and lower visual fields. Topographically, this area is most equivalent to a region in the macaque known alternately as the posterior inferotemporal visual area (PIT, cf. Figs. 6F and 6G) or the temporal-occipital visual area (TEO). This area has been termed V8, but it also has been proposed as an alternative candidate for human V4 due to its selectivity for colored stimuli. The retinotopic organization of this area in both humans and monkeys is less distinct than in more posterior visual areas due to the presence of visually responsive neurons with large receptive fields (at least in the macaque).
Dorsal and lateral to V4v at the junction of the parietal, occipital, and temporal lobes is another visually responsive region that has a representation emphasizing the peripheral visual field (>10°). It is strongly responsive to motion stimuli. This region is likely to be at least partly homologous to the MT-MST complex in macaque monkeys. In the macaque, visual area MT is adjacent to several small, motion-responsive areas (see Fig. 6G). The latter include V4t (a small transition area between V4 and MT), area MST (the medial superior temporal area), and area FST located in the fundus of the superior temporal sulcus. Due to the complexity of this region in the macaque, we prefer to call this region in humans the hMT + complex pending further clarification.
The topography of human visual areas described previously is consistent with earlier anatomical studies of human post mortem material. The latter studies charted the pattern of interhemispheric callosal connections between visual areas in each hemisphere. In monkeys, axons interconnecting visual areas in the left and right hemispheres tend to terminate along the visual field representations of the vertical meridian. Accordingly, they provide an independent source of information about the boundaries of some visual areas, such as V1-V2, VP-V4, and V3-V3A. This is most convincing for the strongly retinotopic areas of visual cortex but becomes increasingly less precise for areas in which the retinotopy breaks down due to the b-
Figure 8 Topography and function of identified visual areas in human occipital lobe and neighboring cortex. Top: Anatomical features and topography of visual areas as displayed on a computer graphics model of the Talairach brain. Whole brain at right shows the plane used to create the separate occipital lobe model. Dashed yellow lines on the model and flat map show a cut along the depths of the calcarine sulcus to permit low-distortion unfolding of the cortical surface. A yellow asterisk marks the tip of the occipital pole. Dark gray outlines on all flat maps represent cortex buried within sulci of the 3-dimensional brain. Scale bar at the far left applies only to the 3-dimensional models. Center: Enlarged flat map of occipital cortex showing the topography of visual areas. Borders between visual areas drawn with solid black lines are relatively well-defined though the exact position of borders will vary from individual to individual. Black arrowheads indicate area borders that correspond with representations of visual field meridia: superior vertical meridian (VMs), inferior vertical meridian (VMi), and horizontal meridian (HM). Visual areas whose borders are shown as dashed lines or shading indicate greater uncertainty at the time of this writing. Center left: Flat maps showing sites (colored dots) reported in the neuroimaging literature to be selectively responsive to some aspect of visual motion (A) or color (C). Each dot is positioned at the point on the surface model that is closest to the Talairach coordinates reported in the original studies. Red-orange dots and swath in the motion map represent the approximate location of hMT + . Purple dots show other motion-responsive sites. Light-green dots and swath represent the location of the kinetic occipital area, KO. Light-blue swath in the color map represents the approximate location of area V8. Center right: Flat maps of form (B) and face or space (D) selective sites. Yellow swath in the form map represents the approximate location of the LO (lateral occipital) visual complex. Dark-green swath in the faces map represents the approximate location of the fusiform face area (FFA), Bottom: Flat maps showing Talairach coordinate isocontours (values in millimeters). Approximate Talairach coordinates of dots in the functional maps can be determined by tracing dots onto a transparent outline of the map and placing it successively over the X, Y, and Z isocontour maps. (See text for a note about accuracy.) Abbreviations: CoS, collateral sulcus; Fusg, fusiform gyrus; IPS, intraparietal sulcus; KO, kinetic occipital visual area; LO, lateral occipital visual complex; LOS, lateral occipital sulcus; MTg, middle temporal gyrus; OTS, occipitotemporal sulcus; POS, parietal-occipital sulcus; STS, superior temporal sulcus; TOS, transverse occipital sulcus. (See color insert in Volume 1).
increasing size of single-cell receptive fields. In these latter areas, the callosal connections become less focused along the vertical meridian and are less useful for marking borders between visual areas. Nevertheless, the results from the human anatomical studies are in reasonably good agreement with the neuroimaging evidence for strongly retinotopic visual areas along the medial and ventromedial wall of the occipital lobe. For visual areas extending onto the dorsolateral and ventrolateral surfaces, the agreement is less certain, in part due to variability in both the imaging and anatomical data within these regions.
Several other visually responsive regions in occipital or nearby parietal and temporal cortex have been proposed as distinct visual areas. Some of these, such as the kinetic occipital area (KO), the lateral occipital area (LO), and the fusiform face area (FFA), are indicated in Fig. 8. Identification of the borders of these areas is more difficult than for the strongly retinotopic areas, and so they are intentionally shown as indistinct in Fig. 8. Other proposed visual areas, such as the parieto-occipital visual area PO-V6 and area V7, are not shown in Fig. 8 because their existence and characteristics are not yet firmly established.
Besides the qualitative aspects of visual field topography used to define visual area boundaries, neuroima-ging has produced new data concerning quantitative features of human retinotopy. Although the qualitative topography of the visual field is preserved in cortical retinotopic maps, the metric of visual space is warped. That is, the representation of the fovea is greatly expanded relative to the representation of the periphery. (In macaques the warping is also known to be different for different visual areas.) Figure 9A shows the representation of visual field eccentricity in striate cortex as described by the researchers Horton and Hoyt in 1991. Their figure was based on a consideration of the extent of blindness in the visual field produced by partial lesions of calcarine occipital cortex. The foveal representation is located posteriorly at the occipital pole, with more eccentric visual field positions (farther from the center of gaze) located successively more anteriorly along the calcarine fissure. The representation of eccentricities close to the center of gaze is expanded relative to more peripheral eccentricities. The representation of the central 10° covers nearly half of the calcarine cortex, with the remainder of the field to beyond 40° covering the other half of cortex. At the time this article was published, little was known of the retinotopic organization of extrastriate cortex in humans. Figures 9B-9D illustrate more recent data from several neuroimaging experiments. As can be seen in Fig. 9C (and Figs. 7D-7F), the representation of visual field eccentricity within V1 of the calcarine sulcus extends both dorsally and ventrally across the entire medial surface of the occipital lobe and onto the dorsolateral and ventral surfaces. As described earlier, this region of cortex contains several different extrastriate visual areas. The results shown in Fig. 9C suggest that the representation of visual field eccentricity in these areas is topographically concordant with the representation in V1. But different visual areas vary in total size, so that the extent of cortex representing 1 ° of visual angle—the cortical magnification factor—will vary from area to area. Cortical mapping functions are illustrated in Figs. 9B and 9D. One study compared
Figure 9 Quantitative retinotopic organization of human visual cortex. (A) Right: Medial view of the occipital lobe with lips of the calcarine sulcus opened to show retinotopy of V1 as estimated from visual field defects in patients with occipital lobe lesions. Numbers along the lip of the calcarine sulcus indicate degrees of visual angle from the center of gaze, as indicated in the adjacent visual field schematic (A, left). (Adapted from Horton and Hoyt, 1991.) (B) Human cortical mapping function according to Engel et al, 1997, Cereb. Cortex 7,181-192. Open symbols are measurements from two observers. The solid curve shows the best fitting exponential (least-squares) to the four hemispheres measured in the study. The dotted line shows an estimate derived from scotomata in human stroke patients and electrophysiological data from nonhuman primates. The x's are fMRI measurements by another researcher (see D). (C) Retinotopy of striate and extrastriate visual cortex as determined by functional MRI displayed on 3-dimensional models of the surface that would correspond to cortical layer 4 (i.e., with supragranular layers removed). The surface model has been smoothed and slightly unfolded to provide a better view of the topography within the calcarine sulcus. White lines represent the loci of constant visual field eccentricity (3, 5, 8, 14, 20, and 24+ degrees of eccentricity). Dashed black lines show representations of horizontal and vertical meridia. Dashed line within the fundus of the calcarine sulcus is the horizontal meridian representation running through the middle of V1. (Based on data from DeYoe et al, 1996, PNAS 93, 2382-2386.) (D) Cortical mapping functions (scale on left axis) and magnification functions (scale on right axis) for upper field striate cortex and three extrastriate visual areas in the human. Also shown are the mapping functions for V1 in the macaque monkey and owl monkey (MM* and OM*, respectively, in the first panel). The cortical mapping functions (heavy lines) show the relationship between distance along the cortex and the representation of visual field eccentricity (Note the difference in graph axis labeling compared to B.) Cortical magnification is the extent of cortex representing 1° visual angle. (Adapted from Sereno et al, 1995, Science 268, 889-893.)
data across visual areas and, for V1, across species. They suggested that there may be an even greater foveal expansion in humans than in monkeys, but this has been challenged. There is some disagreement among studies concerning the degree of expansion of the most central 2-3° in humans. Researchers addressed this issue by comparing their own data, with those of others and with a best fitting exponential function (solid line in Fig. 9B). These studies confirm the contention that the human foveal representation in V1 is larger than had been suggested by earlier studies, and, together, they provide the best quantitative description of visual space mapping in the human visual cortex so far available.
Overall, the topography of occipital visual areas in the human brain is generally consistent with the topography in macaque monkeys, but with some potential departures related to the expansion of the foveal representation and uncertainties concerning areas such as V4 and posterior-lateral occipital cortex. Compared to the macaque monkey, the placement of visual areas relative to anatomical structures in the two species is also different. Human visual areas in lateral cortex appear to be displaced posteriorly toward the occipital pole or onto the medial occipital surface. For example, the hMT+ complex is located posterior to the superior temporal sulcus (STS) in humans but is buried deep within the STS in macaques. Most of the human retinotopic visual areas (V1, V2, V3-VP, V4, V3A) are located entirely or partially on the medial and ventral aspects of the occipital lobe. In macaques, they are located partly or completely on the lateral surface of the brain. This displacement of human visual areas places the foveal representation of the visual field at the occipital pole. In macaques it is situated laterally near the ends of the lunate and inferior occipital sulci. Although the positioning of visual areas relative to anatomical features in humans and macaques is distinctly different, the positioning of visual areas relative to each other may be more consistent across species. This suggests that topographic maps of the visual cortex that are abstracted from gross anatomy (e.g., as seen in cortical flat maps) may provide a more functionally consistent view of cortical organization.
Maps of visual field topography provide one line of evidence for defining and differentiating visual areas. Such maps also provide a good organizational framework for understanding the contributions of different areas to specific visual functions such as color or movement perception. The following section uses this approach to review data concerning the functional specialization of the human visual cortex.
Was this article helpful?