Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 May 8.
Published in final edited form as: Neuron. 2008 May 8;58(3):442–450. doi: 10.1016/j.neuron.2008.02.032

A map for horizontal disparity in monkey V2

Gang Chen 1, Haidong D Lu 1, Anna W Roe 1
PMCID: PMC2441920  NIHMSID: NIHMS50688  PMID: 18466753

Summary

The perception of visual depth is determined by integration of spatial disparities of inputs from the two eyes. Single cells in visual cortex of monkeys are known to respond to specific binocular disparities; however, little is known about their functional organization. We now show, using intrinsic signal optical imaging and single unit physiology, that, in the thick stripe compartments of the second visual area (V2), there is a clustered organization of Near cells and Far cells and, moreover, there are topographic maps for Near to Far disparities within V2. Our findings suggest that maps for visual disparity are calculated in V2, and demonstrate parallels in functional organization between the thin, pale, and thick stripes of V2.

Introduction

Binocular inputs are used by the human visual system to judge object depth in the three dimensional world. Stimuli nearer or farther from the fixation point will produce disparities from left and right eye with a negative or positive horizontal shift, respectively. Neurons selective for horizontal disparities are first observed in primary visual cortex (V1) and have been characterized with different methodologies. These include response to offset bars presented to the two eyes (‘tuned excitatory’, ’tuned inhibitory’, ‘near’, and ‘far’ cells, Poggio and Fischer, 1977), response to differential phase of sinusoidal gratings (Ohzawa and Freeman, 1986), and response to absolute disparity of random dot stereograms (Cumming and Parker, 1997). In the second visual cortex (V2), disparity-selective neurons have been described as ‘obligatory binocular’ (Hubel and Wiesel, 1970), selective for disparity-defined contours (Bakin et al., 2000; Qiu and von der Heydt, 2005), and tuned for relative disparity (Thomas et al., 2002). Despite the number of studies on disparity responses in the visual cortex, there are no published studies addressing any systematic representation of disparity response in either V1 or V2 (Burkitt et al., unpublished results; Cumming and DeAngelis, 2001).

Of the three featural maps represented in V2, two have been shown to exhibit a topographic organization. Functional maps of contour orientation exhibit systematic orientation structure characterized by pinwheels in the thick and pale stripes (Malach et al., 1994; Ramsden et al., 2001; Ts'o et al., unpublished results; Xu et al., 2004a). In the thin stripes, color representation in V2 exhibits a systematic topography for hue representation (Xiao et al., 2003) as well as an organization for brightness increment (ON) vs. decrement (OFF) representation (Roe et al., 2005b; Wang et al., 2006). Given the known organizations of the other featural maps in V2 (color and contour in the thin and pale/thick stripes of V2, respectively), this gap in our knowledge regarding disparity representation in V2 is particularly conspicuous. Disparity-selective responses are believed to be preferentially localized to the cytochrome oxidase thick stripes of V2, but topographic representation within these stripes has not been examined (Bakin et al., 2000; Burkitt et al., unpublished results; Hubel and Livingstone, 1987; Peterhans and von der Heydt, 1993; Roe and Ts'o, 1995).

Thus, at issue are the questions of whether there is a map for Near to Far depth information and whether disparity and orientation information, which are both represented in the thick stripes, are independently represented or not. The broader issue of parallelism and modularity across featural domains in V2 is also in question. Examination of these issues is further motivated by the presence of maps organized for disparity in area MT (DeAngelis and Newsome, 1999), a primary target of V2 thick stripes.

Results

To address these questions, we used intrinsic signal optical imaging methodology (Grinvald et al., 1986) and single unit electrophysiology. Nine hemispheres from six anesthetized monkeys were studied. Eyes were stabilized and prisms were used to position the view of each eye on a monitor. Spot imaging and electrophysiological methods were used to determine the visual extent of the imaging field of view and the precise convergence of the two eyes (Supplementary Figures S1 & S2; Lu et al., submitted ). Disparity stimuli consisted of random dot stereograms (random dot textures presented with a relative shift of the dots in one eye relative to the other, Julesz, 1971) presented at different ocular offsets. These stimuli have the advantage that they induce strong percepts of depth (Near or Far depending on shift between the two eyes) purely by interocular offset without any dependence on contour orientation, spatial contrast, or spatial frequency, complications of both bar and grating stimuli. Such random dot stereograms (RDS) were presented with seven or thirteen different horizontal disparities to induce percepts of surfaces at depth ranging from 0.85° near to 0.85° far (see Methods). As the extent of the perceived square (6°×6°) induced by the RDS was outside the imaged field of view, the imaged signal can be attributed to RDS-induced depth uncontaminated by possible depth-related illusory contour signals (Qiu and von der Heydt, 2005). Uncorrelated stereograms, which contain similar monocular cues as RDS, failed to induce constant depth surface percepts and were used as control comparisons (Gonzalez and Perez, 1998).

Preferential activation in V2 thick stripes

Using standard optical imaging techniques, we mapped the locations of the V1/V2 border (Figure 1D), and the thin, pale, and thick stripes in V2 (Blasdel, 1992; Roe et al., 2005a; Ts'o et al., unpublished results; Xu et al., 2004a). As confirmed with cytochrome oxidase histology (Figures 1A and 1B) and consistent with previous studies, orientation domains overlay thick and pale stripes (Figure 1E), domains for color vs. luminance preference overlay thin stripes (Figure 1F), and domains for binocularity vs. monocularity overlay the thick stripes (Figure 1G).

Figure 1. Localization of the disparity response in the thick stripes of V2.

Figure 1

Open in a new tab

(A) A section stained for cytochrome oxidase showing the positions of thin (red arrow heads) and thick (a black arrow head) stripes of V2. (B) Enlarged part in red box of (A). (C) Percentage of pixels in V1 and in the thin, pale, and thick stripes of V2 with significant (p < 0.05) response to RDS. More pixels in thick stripes (n = 5) responded to depth stimuli than in the pale or thin stripes within V2 or in V1 (p < 10−6, t test). Error bars, ± SEM. (D) Ocular dominance map in V1 (left eye minus right eye) reveals V1/V2 border (horizontal line). (E) Orientation map (differential response, horizontal vs. vertical gratings). (F) Thin stripes are determined by the areas which prefer color to luminance. Note that regions with color preference (thin stripes) have poor orientation preference. The gray region is out of field of view as camera was moved to slightly different location for the color run. (G) Difference between binocular stimuli (dark pixels) and monocular stimuli (light pixels) reveal the position of the thick stripe (cf. Ts’o et al 2001). (H) RDS disparity map. Sum of all Near and Far. (I) White pixels are those with significant responses to RDS (compared to responses to uncorrelated stereograms, t test, p < 0.05). Scale bar: 1 mm (A) and (B). Scale bar for (B) applies to (D)–(I). A, anterior; M, medial.

We found that the areas with significant responses to RDS (based on pixel-by-pixel t test of RDS and uncorrelated stereograms, p < 0.05) were preferentially located in the thick stripes of V2 (Figures. 1H and 1I). On average, roughly half of the pixels (43.1% ± 3.8) within thick stripes (n = 9) were significantly responsive to at least one of the RDS defined surfaces (Figure 1C). In comparison, a much lower fraction of the pixels in V1 (8.2% ± 1.2, n = 9), V2 thin stripes (10.6% ± 1.3, n = 9), and V2 pale stripes (14.5% ± 1.5, n = 9) exhibited significant activation to RDS (Figure 1C). These data suggest a preferential response to RDS in V2 thick stripes that is not present elsewhere in V2 or in V1. In sum, disparity selective cells in V2 thick stripes exhibit preferential binocular vs. monocular response (Ts'o et al., 2001), responsiveness to depth contours (Bakin et al., 2000; Qiu and von der Heydt, 2005), and surface depth induced by RDS (Thomas et al., 2002). Such organized responses are notably absent from the other stripes of V2 and from V1.

A map for horizontal disparity in V2

To examine possible segregation of Near and Far responses in the thick stripes, we computed a Near minus Far difference map (Figure 2A, obtained by taking the difference of the sums of the three Near conditions and the three Far conditions). In both thick stripes visible in this field of view, this map reveals a clear segregation of Near preference (dark pixels) and Far preference (light pixels). Interestingly, the mediolateral position of Near vs. Far appears to differ in these two stripes.

Figure 2. Disparity topography in V2.

Figure 2

Open in a new tab

(A) Differential image between all Near (dark pixels) and all Far (light pixels) stimuli. Black dashed line, border of V1–V2. (B)–(H). Single condition images evoked by three Near stimuli with disparities of 0.34°(B), 0.17°(C), and 0.085°(D), Zero (E), and three Far stimuli of 0.085°(F), 0.17°(G) and 0.34°(H). Positions of two crosses are constant through (A)–(I). The outlined regions in (B)–(H) show areas with significantly greater response to RDS than uncorrelated stereograms (p < 0.05, t test). As described by ellipse fits, the lengths of these domains in (B)–(H) are 802, 831, 856, 933, 1216, 690, and 918 µm, respectively. Widths are 431, 428, 350, 437, 370, 549, and 298µm, respectively. (I) Summary showing overlay of three disparity contours (red: 0.34° near, green: zero, blue: 0.34° far). (J) Optical image of response to uncorrelated random dots (vs. Blank. A streak of white activity (upper right corner) is due to blood vessel noise). Grayscale: magnitude of imaged response in percent reflectance. Color scale: near to far disparity. Scale bars in A, B: 1 mm. A, anterior; L, lateral.

Further examination of single condition maps revealed a clear Near-to-Far topography. Figures 2B–H illustrate single condition maps of one thick stripe (indicated by the dotted box in Figure 2A) in response to each of seven RDS disparity stimuli: Near 0.34°, 0.17°, 0.085°, Zero (Figures 2B–E), and Far 0.085°, 0.17°, and 0.34° (Figures 2F–H). In each panel, the region significantly activated by the RDS stimulus (p < 0.05, t test) is outlined in color and appears as an elongated band, roughly 200 µm in width and 1 mm in length. As the stimulus disparity shifts from Near to Far (Figures 2B–H), the activation band shifts from left to right (medial to lateral). This topography is summarized in Figure 2I (3 representative disparities: −0.34° red, 0° green, and +0.34° blue). In comparison, responses to uncorrelated random dot stereograms failed to produce any structured responses (Figure 2J).

A second case is shown in Figure 3. In this case, random dot stereograms elicited multiple foci of activation within a single thick stripe. As horizontal disparity shifted from −0.34° (B), −0.17° (C), −0.085° (D), zero (E), +0.085° (F), +0.17° (G), and +0.34° (H), the location of each of these domains also shifted. If one follows a single activation site (labeled 1, 2, 3 in Figure 3B) a consistent topography for each site is still evident (shifting outlines shown in Figures 3I–K). Three more cases are shown in supplementary materials (Supplementary Figure S3), which include two cases (Case 2 and Case 3) examining a larger disparity range from +0.85° Far to −0.85° Near.

Figure 3. Disparity topography in V2: a second case.

Figure 3

Open in a new tab

(A) Differential image between all Near (dark pixels) and all Far (light pixels) stimuli. Dotted box: region shown in B–K. Black dashed line: V1/V2 border. (B–H), Single condition disparity images: −0.34°(B), −0.17°(C), −0.085°(D), zero (E), +0.085°(F), +0.17°(G), and +0.34°(H). Three activation domains are labeled 1, 2, and 3 in B. Color outlines: regions of significant activation (p < 0.05, t test). (I–K), Summary topographies of domains 1 (I), 2 (J), and 3 (K). Conventions same as Figure 2. Scale bars: 1 mm. A, anterior; L, lateral.

To quantify this shift, we defined two indices, one to reflect percentage overlap and another to reflect the distance between the centers of different domains (see Methods). If representations of disparity in V2 were topographic, then one would predict that domains representing similar disparities would be closer together and exhibit more overlap, while those representing different disparities would be further apart and exhibit less overlap. Figures 4A and 4B support these predictions. As shown in Figure 4A, the overlap ranged from more than 50% for the smallest disparity differences (Δ0.085°, 52.5 ± 2.8 %, n = 140) to almost no overlap at disparity differences of more than half a degree (Δ0.68°, 0.2 ± 0.2 %, n = 30, Δ0.60°, Δ0.77°, Δ0.85°, Δ0.94°, and Δ1.02° no overlap). Delta disparity correlates significantly with overlap (p < 10−10, ANOVA; correlation regression, r = 0.42, p < 0.01, n = 708); on average, a 1° disparity shift results in a 77.9% decrease in overlap. Similarly, the distance between domains (center of mass) varied with delta disparity (p < 10−10, ANOVA, Figure 4B). Over the disparity range tested, the distance between centers increased from 180 µm (Δ0.085°, 181.6 ± 21.5 µm, n = 70) to almost four times greater (Δ0.65°, 699.6 ± 153.4 µm, n = 4) with increasing disparity. These tendencies were found in all six animals we tested (Supplementary Figure S4).

Figure 4. Overlap and distance measurements support topography.

Figure 4

Open in a new tab

(A) Percent overlap between pairs of disparity domains. Percent overlap predicts delta disparity: the greater the overlap the more similar the disparity. (B) Distance between centers of mass of disparity domains. Distance predicts delta disparity: smaller distances predict similar disparity. Numbers of domain pairs are indicated by numbers in bars. Since no two disparity domains have Δdisparity larger than 1.1°, the largest bin shown is 1.1° of disparity difference. (C–E), average length, width, and length-width ratio of All (black, n = 123), Near (red, n = 45), Zero (green, n = 50), Far (blue, n = 28) from 9 cases. Error bars, ± SEM.

Cortical magnification is an important aspect of any topographic map. The possibility that there exist differential cortical magnifications across a parameter space can indicate the relative importance or precision of that parameter within certain ranges of that space. We therefore considered the possibility that magnifications are greater or smaller within the range of depths examined. However, when the size and shape of the activated bands were measured by fitting the iso-disparity regions with ellipses (Figures 4C–E), we found that on average the length and width of Near domains (red, n = 45), Zero domains (green, n = 50), and Far domains (blue, n = 28), did not differ from each other (both p > 0.5, ANOVA). Based on this roughly even representation within the −0.5° to +0.5° disparity and the measurements in Figure 4A and 4B, the cortical magnification factor of disparities is on average 1° per 780 µm.

Neural basis

To examine the neuronal basis of the optical signal, we targeted imaged disparity domains with microelectrodes and recorded single unit activity from the superficial cortical layers (Figure 5A, same field view as Figures 2B–J). In each penetration, 1–4 units were isolated and disparity tuning curves obtained. Within each penetration, cells shared similar tuning preferences (Figure 5B), consistent with previous recordings within vertical penetrations (Ts'o et al., 2001). We then normalized and averaged disparity tuning curves recorded in each penetration. Plotted in Figure 5D is the average neuronal disparity tuning curve for each of the four penetrations shown in Figure 5A. We then compared this neural tuning curve to the optical responses at each recording location (Figure 5C). Comparison of Figure 5C and 5D reveal that the neuronal and optical response at these four locations had similar tuning curves (Figure 5E, robust fit, r = 0.80, p < 0.01, n = 28). In fact, for the entire population of responses we recorded (individual tuning curves shown in Supplementary Figure S5), the average neuronal disparity response preference showed significant correlation with optical disparity response preference (Figure 5F, robust fit, r = 0.84, p < 0.01, n = 27). A similar tendency was found for disparity tuning width (Figure 5G, robust fit, r = 0.73, p < 0.01, n = 27). Deviations from a perfect correlation and tuning width extracted from optical signal can be at least partly attributed to the fact that the optical signal derives from a population average of neuronal responses. Thus, the neuronal responses are consistent with the presence of topography for disparity in V2 and indicate the presence of neurons that respond to RDS.

Figure 5. Comparison between neuronal and optical signals.

Figure 5

Open in a new tab

(A) Four electrode-recording locations (marked by colored squares) within the color-coded disparity domains (see color scale bar). (B) Example consistent with columnar organization. Disparity tuning of four neurons recorded within a single vertical penetration. Red lines: Gaussian fit. Blue triangles: preferred disparity. Normalized optical (C) and neural (D) disparity tuning curves at each of 4 locations in A. Dashed vertical lines in C–D: zero disparity. (E), scatter plot for each pair of optical (abscissa) and neuronal (ordinate) signals in C–D, showing significant correlation (robust regression, r = 0.80, p < 0.01). Scatter plot for disparity preference (F) and disparity tuning width (G) obtained from optical and neuronal signals from 3 cases did both tests (n = 27). Disparity preferences are similar (p > 0.4, paired t test) while tuning widths tend to be wider than optical signals (see histograms at upper right corners, p < 0.01, paired t test). Different marker shapes are neurons recorded from different monkeys. For detailed optical and neuronal tuning curves, see Supplementary materials (Supplementary Figure S5). Thick lines are diagonals. Scale bars: 1 mm (A), 0.25° (D). Error bars, ± SEM.

Relationship to orientation

A central issue in cortical representation is how multiple maps co-exist within the same two-dimensional cortical space. In V1, many modeling studies suggest that competing intracortical influences result in observed functional organizations for ocular dominance and orientation (Swindale, 2004). In V2, studies suggest an interdigitation of different featural maps that parallel the stripe-like organization revealed in cytochrome oxidase stains (Roe and Ts'o, 1995; Swindale, 2004). However, the relative organization of orientation and disparity in V2 has been elusive. Electrophysiological evidence suggest that representation of horizontal disparity might be biased for vertical orientations (Cumming, 2002: the frequency of cells with horizontal orientation preference is about one third of that with vertical orientation preference in V1; Hubel and Wiesel, 1970). This suggests a possible overrepresentation of vertical orientation relative to horizontal orientation within the V2 thick stripes.

To investigate this co-representation of orientation and disparity within V2, we examined representation of disparity within orientation domains and the representation of orientation within disparity domains. In this analysis, we considered only the pixels which exhibited significant disparity response (Figure 6A). We found that the disparity activated regions contained roughly equal representation for different orientations (Figure 6B, p > 0.9, ANOVA, n = 9) and that on average each orientation domain overlay a range of disparity domains (Figure 6C, no significant influence from disparity, orientation preferences, and between them to the overlay, all p > 0.8, balanced two-way ANOVA, n = 9). Thus, we found no evidence for increased or biased disparity representation in vertical orientation domains. Furthermore, we tested whether or not the sensitivities to disparity are similar among different orientation domains. We compared the disparity tuning strength within different orientation domains and found the horizontal orientation domains are as sharply tuned for disparity as neurons preferred other orientations (Figure 6D, p > 0.9, ANOVA, n = 9). Our data suggest, at least for the disparity ranges tested here, orthogonality between the disparity and orientation parameters within the V2 thick stripes (Figure 7).

Figure 6. Relationship between orientation domains and disparity domains.

Figure 6

Open in a new tab

(A) The orientation selectivity of areas with significant response to any disparity tested (same case as shown in Figure 1). The preferred orientations are shown (color wheel). (B) Percentage of pixels (of the disparity selective pixels) at each of the 0°, 45°, 90°, and 135° orientations. Within disparity selective pixels, across all cases, the average percentage of areas covered by each orientation was not significantly different from 25% (all p > 0.5, t test). (C) Statistical results of the average percentage of pixels in an orientation domain covered by different disparity domains. Based on a balanced two-ways ANOVA, where the factors were disparity, orientation, and their interaction, there is no significant influence from orientation or from disparity (both p > 0.8) or between them (p > 0.9). (D) Averaged disparity tuning strength within orientation preference domains. No significant difference was found between horizontal and 45°, 90°, or 135° orientation preference domains (all p > 0.5, t test, n = 9). Error bars, ± SEM. Scale bar: 1 mm (A).

Figure 7.

Figure 7

Open in a new tab

Depiction of possible organization in V2 where orientation and disparity parameters are orthogonally represented.

Discussion

Disparity response in V2

The role of V2 in depth perception has long been hypothesized. Disparity selective cells, especially those termed ‘obligatory binocular’ are prevalent in V2 (Baizer et al., 1977; Hubel and Wiesel, 1970). The role of V2 in depth perception is further supported by the presence of cells that detect disparity-defined contours and disparity capture response (Bakin et al., 2000; Qiu and von der Heydt, 2005), response to relative over absolute disparity (Thomas et al., 2002), and stereopsis choice related activity (Nienborg and Cumming, 2006, 2007). Functional MRI studies have also found significant disparity-induced activations within V2 (Sereno et al., 2002; Tsao et al., 2003). The preferential, though not exclusive, localization of disparity response to the thick stripes of V2 has been demonstrated electrophysiologically (Bakin et al., 2000; Hubel and Livingstone, 1987; Peterhans and von der Heydt, 1993) and with optical imaging methods which localized preferential response to two eyes over one eye (Ts'o et al., 2001).

Disparity topography in V2

Although the functional organization of V2 thick stripes was poorly understood, it was suggested that disparity, like color and contour, is represented in a modular fashion (Hubel and Livingstone, 1987; Roe, 2003; Roe and Ts'o, 1995; Ts'o et al., 2001). The possible presence of clustered organization is one indicator of organized topography. Ts'o et al (2001) reported that the orientation of obligatory binocular cells in vertical penetrations tend to be similar, although they find disparity tuning shift slightly with depth, an observation that may be related to slightly off-orthogonal penetrations. Evidence for clustered response in V2 is also reported by Nienborg and Cumming (2006) who find preferred disparities of multi- and single-unit recordings from the same location are similar. A similar but weaker tendency was found in V1 (Prince et al., 2002). Our small sample of electrophysiological recordings is also consistent with the presence of columnar organization (Figure 5B).

With optical imaging it is possible to observe topographic organizations at the sub-millimeter scale. In this study, we found that horizontal disparity is mapped in a topographic fashion in V2. Domains were on the order of 200 µm in width and mapped in a systematic fashion within thick stripes. In some cases, multiple topographic disparity maps were observed within V2 thick stripes (Figure 3). It is possible these multiple domains could represent different visuotopic locations in space. Alternatively, different disparity-selective clusters could represent different functionalities which further differentiate different types of disparity response (e.g. those related to motion borders, Lu et al., unpublished data). Such functional distinctions would parallel those found in other stripe types (e.g. hue vs. on/off domains within V2 thin stripes). This is an interesting issue that remains to be investigated.

The presence of topographic organization within the thick stripes is consistent with the presence of featural maps for color and contour in thin and pale stripes, respectively. Functional maps of contour orientation, both real contour response and illusory contour response, exhibit systematic orientation structure characterized by pinwheels in the thick and pale stripes (Roe and Ts'o, 1995; Ts'o et al., 2001). In the thin stripes, color representation in V2 exhibits a systematic topography for hue representation (Xiao et al., 2003) as well as an organization for brightness increment vs. decrement representation (Wang et al., 2006). Our findings extend this organization to the thick stripes and a consistent framework for understanding V2 function.

Relationship between disparity and orientation response

We found that orientation and horizontal disparity are mapped somewhat independently. Although some studies report a predominance of disparity-selective preferring vertical orientations (Hubel and Wiesel, 1970; Poggio and Fischer, 1977), our study finds no such preferential association. Indeed, disparity responsive regions contained orientation response for all orientations tested in roughly equal proportions and orientation domains contained a range of disparity response. However, the possibility remains that the disparity selective response evoked by a 6.0°×6.0° region in depth may stimulate a different population of cells than those responsive to contour disparity.

Disparity organizations in other areas

Other cortical areas in primate besides V2 also exhibit functional architecture for horizontal disparity. In macaque MT, disparity tuned neurons are organized in a columnar fashion and, within disparity preferring regions, exhibit a smoothly shifting topographic organization for disparity selectivity (DeAngelis et al., 1998; DeAngelis and Newsome, 1999). In V3 (Adams and Zeki, 2001) and V4 (Watanabe et al., 2002), neurons with similar disparity selectivity are clustered. Input from V2 might play an important role in either forming or conferring organization in these areas. V3 has major connections with area V2 (Felleman et al., 1997; Stepniewska and Kaas, 1996) and V4 receives inputs both from V2 and from V3 (Stepniewska and Kaas, 1996; Zeki and Shipp, 1989). In particular, as thick stripes of V2, unlike the thin or pale stripes, project directly to MT (DeYoe and Van Essen, 1985; Shipp and Zeki, 1985, 1989), this raises the possibility that the systematic organization of disparity in MT is initiated in V2. This possibility is further strengthened by the finding that inactivation of V2/V3 attenuates the sensitivity of MT neurons to binocular disparity (Ponce et al., 2008). Thus, V2 may be an important bridge both for generalizing depth-related activity from primary visual cortex (Bakin et al., 2000; Cumming and Parker, 1999; Thomas et al., 2002) and for conferring organized depth-related information to higher cortical areas.

In summary, this study demonstrates topography for disparity in V2. Localization of this topography to the thick stripes of V2 further suggests that modular organization is a feature shared across different stripe types and different featural spaces in V2. The additional constraints provided by these data will be useful for models of functional organization within and beyond V1. These data also raise the possibility that at least some aspects of the disparity organizations observed in MT derive from V2 and are not established de novo in MT.

Experimental Procedures

Visual stimulus and Data collection

Experiments were preformed under protocols conforming to guidelines of the National Institutes of Health and approved by Vanderbilt University Animal Care and Use Committees. Detailed procedures are described elsewhere (Roe et al., 2005b). Briefly, macaque monkeys were anesthetized (i.v., thiopental sodium, 1–2 mg·kg−1·h−1), paralyzed with vercuronium bromide (i.v., 100 µg·kg−1·h−1), and artificially ventilated. Anesthetic depth was assessed continuously by EEG, end-tidal CO2, pulse oximetry, and heart rate. Eyes were dilated (atropine sulfate) and fitted with contact lenses to focus on a monitor 76.2 cm from the eyes. Craniotomy and durotomy were performed to expose visual areas V1 and V2.

Visual stimuli were generated with custom software and were presented at 100 Hz. Each eye was stimulated independently, by diverging eyes with Risley prisms to different parts of the screen. Random dot stereograms which produce constant depth surface percepts (50% dark (0.0 cd·m−2)/50 % bright (80.0 cd·m−2) dots, dot size 0.085°×0.085°, dot density 100%) were presented on a gray background (40.0 cd·m−2). A new dot pattern was presented every 100 ms. Horizontal binocular disparities were introduced by shifting the location of dots in the two eyes. The stereograms consisted of an 8.5°×8.5° square background region maintained at zero disparity, and a 6.0°×6.0° center patch whose disparity was varied from trial to trial. The tested disparity levels ranged between −0.85° (crossed) to +0.85° (uncrossed). In six hemispheres, we tested seven different disparity levels (three Near: 0.34°, 0.17°, 0.085°, Zero: 0°, and three Far: 0.085°, 0.17°, 0.34°) during imaging session. Single units were tested from −0.425° to +0.425° with steps of 0.085°. A wider range of disparities were tested on the remaining three hemispheres (13 levels, six Near: 0.85°, 0.68°, 0.51°, 0.34°, 0.17°, 0.085°, Zero: 0°, and six Far: 0.085°, 0.17°, 0.34°, 0.51°, 0.68°, 0.85°) during imaging sessions and single units were tested from −0.85° to +0.85° with steps of 0.17°. Uncorrelated stereograms, which do not elicit a constant depth surface percept, were used as controls. Full screen square gratings (spatial resolution: 1 cycle/°) were used to test the orientation preference of the visual cortex.

To eliminate possible ocular drift, eyes were stabilized with eye rings. In addition, eye position was repeatedly checked throughout the experiment by a rapid, high spatial precision spot imaging method (Supplementary Figures S1 & S2, Lu et al., submitted). Rapid spot imaging was also used to confirm that the imaged field of view lay well within the bounds the RDS depth square.

Images of cortical reflectance change (intrinsic signals) were acquired by using IMAGER 3001 (Optical Imaging, Germantown, NY) and 630 nm illumination. For each condition, 30 trials were presented. Areas containing extensive vascular artifact were excluded from analysis.

Data analysis

In our standard optical imaging analysis, signal-to-noise ratio was enhanced by trial averaging. Images were smoothed by a disc mean filter kernel (radius 80 µm). Low frequency noise was reduced by convolving the image with an 800 µm radius mean filter kernel and subtracting the result from the original image (Xu et al., 2004b). Single condition disparity maps (Figure 2 and Figure 3) were obtained by summing frames over 3 sec and subtracting control maps. As, across trials, the pixel response is normally distributed (Lilliefors test, 93.4 ± 0.2% pixels with p > 0.05, n = 12), we used t test to test the significance. Areas (larger than 100 µm in diameter) were considered to have significant activation (p < 0.05) when on average the response of optical signals to RDS are greater than those to uncorrelated stereograms controls based on a trial-trial comparison. Activation zones were fitted with an ellipse (Fitzgibbon et al., 1999). The following indices were calculated:

Center of mass:[COMxCOMy]=[1n∑i=1nxi1n∑i=1nyi].
Distance between domains:DCOM=(COMx,1−COMx,2)2+(COMy,1−COMy,2)2.
Overlap between domains A and B:POOA,B=AreaA∩BAreaA×100

AreaA∩B is size of the area common to A and B.

Disparity tuning strength: S=|Rmax−Rmin||Rmax−Rmin|+2×SSEN−M was defined in a manner similar as disparity discrimination index (DeAngelis and Uka, 2003; Hinkle and Conner, 2005; Prince et al, 2002), where SSE is the sum squared error around the mean responses, N is the total number of stimulus presentation, M is the number of different disparities tested, Rmax and Rmin represent the mean responses to the most effective and least effective stimuli, respectively. This index lies between 0 and 1. A larger index indicates more sharply tuned disparity.

The color-coded orientation preference map was derived from pixel-by-pixel vector summation (Blasdel, 1992; Bonhoeffer and Grinvald, 1991; Bosking et al., 1997). Two tailed t test was applied except where indicated.

Electrical recording

For each recording, two single neurons were isolated, one in V1 (a binocular cell used as a reference cell (Anzai et al., 1999; Ferster, 1981; Hubel and Wiesel, 1970; Roe and Ts'o, 1995) and one in V2. For each V2 cell isolated, the center of receptive field was plotted and the RDS centered on the receptive field of each eye. Different horizontal disparities were presented for 2 second each. A gray screen was presented during the 2 second inter-stimulus interval. RDS conditions were interlaced in pseudo-random fashion (3–10 repetitions each). Disparity tuning curves were fit with Gaussians by robust regression. For cells with tuned inhibitory tuning curves, negative-going Gaussians were used. For average tuning curves, responses to different disparities were normalized. Error bars are ± SEM.

Histology

At the end of experiments, animals were given a lethal dose of sodium pentothal. Cytochrome-stained tissue was aligned with optical images using lesion sites, cytochrome oxidase landmarks, and blood vessels as guides (Roe et al., 2005a; Shmuel et al., 2005; Xu et al., 2004a).

Supplementary Material

01

Acknowledgements

We thank Drs F. Tong and V. Casagrande for suggestions about this manuscript and Drs. Y. Kamitani, and X. Zha for discussion, Peter Kasken for histology, Daniel Shima for programming, Yanyan Chu for technical assistance, and Dr. Jon Kaas for providing eye rings. This work was supported by NIH grant EY 11744 to A. W. R, Vanderbilt Vision Research Center, and Vanderbilt University Center for Integrative & Cognitive Neuroscience.

Footnotes

Author Information The authors declare no competing financial interests.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Adams DL, Zeki S. Functional organization of macaque V3 for stereoscopic depth. J Neurophysiol. 2001;86:2195–2203. doi: 10.1152/jn.2001.86.5.2195. [DOI] [PubMed] [Google Scholar]
  2. Anzai A, Ohzawa I, Freeman RD. Neural mechanisms for encoding binocular disparity: receptive field position versus phase. J Neurophysiol. 1999;82:874–890. doi: 10.1152/jn.1999.82.2.874. [DOI] [PubMed] [Google Scholar]
  3. Baizer JS, Robinson DL, Dow BM. Visual responses of area 18 neurons in awake, behaving monkey. J Neurophysiol. 1977;40:1024–1037. doi: 10.1152/jn.1977.40.5.1024. [DOI] [PubMed] [Google Scholar]
  4. Bakin JS, Nakayama K, Gilbert CD. Visual responses in monkey areas V1 and V2 to three-dimensional surface configurations. J Neurosci. 2000;20:8188–8198. doi: 10.1523/JNEUROSCI.20-21-08188.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Blasdel GG. Orientation selectivity, preference, and continuity in monkey striate cortex. J Neurosci. 1992;12:3139–3161. doi: 10.1523/JNEUROSCI.12-08-03139.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bonhoeffer T, Grinvald A. Iso-orientation domains in cat visual cortex are arranged in pinwheel-like patterns. Nature. 1991;353:429–431. doi: 10.1038/353429a0. [DOI] [PubMed] [Google Scholar]
  7. Bosking WH, Zhang Y, Schofield B, Fitzpatrick D. Orientation selectivity and the arrangement of horizontal connections in tree shrew striate cortex. J Neurosci. 1997;17:2112–2127. doi: 10.1523/JNEUROSCI.17-06-02112.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cumming BG. An unexpected specialization for horizontal disparity in primate primary visual cortex. Nature. 2002;418:633–636. doi: 10.1038/nature00909. [DOI] [PubMed] [Google Scholar]
  9. Cumming BG, DeAngelis GC. The physiology of stereopsis. Annu Rev Neurosci. 2001;24:203–238. doi: 10.1146/annurev.neuro.24.1.203. [DOI] [PubMed] [Google Scholar]
  10. Cumming BG, Parker AJ. Responses of primary visual cortical neurons to binocular disparity without depth perception. Nature. 1997;389:280–283. doi: 10.1038/38487. [DOI] [PubMed] [Google Scholar]
  11. Cumming BG, Parker AJ. Binocular neurons in V1 of awake monkeys are selective for absolute, not relative, disparity. J Neurosci. 1999;19:5602–5618. doi: 10.1523/JNEUROSCI.19-13-05602.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. DeAngelis GC, Cumming BG, Newsome WT. Cortical area MT and the perception of stereoscopic depth. Nature. 1998;394:677–680. doi: 10.1038/29299. [DOI] [PubMed] [Google Scholar]
  13. DeAngelis GC, Newsome WT. Organization of disparity-selective neurons in macaque area MT. J Neurosci. 1999;19:1398–1415. doi: 10.1523/JNEUROSCI.19-04-01398.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. DeAngelis GC, Uka T. Coding of horizontal disparity and velocity by MT neurons in the alert macaque. J Neurophsiol. 2003;89:1094–1111. doi: 10.1152/jn.00717.2002. [DOI] [PubMed] [Google Scholar]
  15. DeYoe EA, Van Essen DC. Segregation of efferent connections and receptive field properties in visual area V2 of the macaque. Nature. 1985;317:58–61. doi: 10.1038/317058a0. [DOI] [PubMed] [Google Scholar]
  16. Felleman DJ, Burkhalter A, Van Essen DC. Cortical connections of areas V3 and VP of macaque monkey extrastriate visual cortex. J Comp Neurol. 1997;379:21–47. doi: 10.1002/(sici)1096-9861(19970303)379:1<21::aid-cne3>3.0.co;2-k. [DOI] [PubMed] [Google Scholar]
  17. Ferster D. A comparison of binocular depth mechanisms in areas 17 and 18 of the cat visual cortex. J Physiol. 1981;311:623–655. doi: 10.1113/jphysiol.1981.sp013608. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Fitzgibbon A, Pilu M, Fisher RB. Direct Least Square Fitting of Ellipses. IEEE Transactions Pattern Analysis and Machine Intelligence. 1999;21:476. [Google Scholar]
  19. Gonzalez F, Perez R. Neural mechanisms underlying stereoscopic vision. Prog Neurobiol. 1998;55:191–224. doi: 10.1016/s0301-0082(98)00012-4. [DOI] [PubMed] [Google Scholar]
  20. Grinvald A, Lieke E, Frostig RD, Gilbert CD, Wiesel TN. Functional architecture of cortex revealed by optical imaging of intrinsic signals. Nature. 1986;324:361–364. doi: 10.1038/324361a0. [DOI] [PubMed] [Google Scholar]
  21. Hinkle DA, Connor CE. Quantitative characterization of disparity tuning in ventral pathway area V4. J Neurophysiol. 2005;94:2726–2737. doi: 10.1152/jn.00341.2005. [DOI] [PubMed] [Google Scholar]
  22. Hubel DH, Livingstone MS. Segregation of form, color, and stereopsis in primate area 18. J Neurosci. 1987;7:3378–3415. doi: 10.1523/JNEUROSCI.07-11-03378.1987. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hubel DH, Wiesel TN. Stereoscopic vision in macaque monkey. Cells sensitive to binocular depth in area 18 of the macaque monkey cortex. Nature. 1970;225:41–42. doi: 10.1038/225041a0. [DOI] [PubMed] [Google Scholar]
  24. Julesz B. Foundations of Cyclopean Perception. Chicago: The University of Chicago Press; 1971. [Google Scholar]
  25. Malach R, Tootell RB, Malonek D. Relationship between orientation domains, cytochrome oxidase stripes, and intrinsic horizontal connections in squirrel monkey area V2. Cereb Cortex. 1994;4:151–165. doi: 10.1093/cercor/4.2.151. [DOI] [PubMed] [Google Scholar]
  26. Nienborg H, Cumming BG. Macaque V2 neurons, but not V1 neurons, show choice-related activity. J Neurosci. 2006;26:9567–9578. doi: 10.1523/JNEUROSCI.2256-06.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Nienborg H, Cumming BG. Psychophysically measured task strategy for disparity discrimination is reflected in V2 neurons. Nat Neurosci. 2007;10:1608–1614. doi: 10.1038/nn1991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Ohzawa I, DeAngelis GC, Freeman RD. Stereoscopic depth discrimination in the visual cortex: neurons ideally suited as disparity detectors. Science. 1990;249:1037–1041. doi: 10.1126/science.2396096. [DOI] [PubMed] [Google Scholar]
  29. Ohzawa I, Freeman RD. The binocular organization of simple cells in the cat's visual cortex. J Neurophysiol. 1986;56:221–242. doi: 10.1152/jn.1986.56.1.221. [DOI] [PubMed] [Google Scholar]
  30. Peterhans E, von der Heydt R. Functional organization of area V2 in the alert macaque. Eur J Neurosci. 1993;5:509–524. doi: 10.1111/j.1460-9568.1993.tb00517.x. [DOI] [PubMed] [Google Scholar]
  31. Poggio GF, Fischer B. Binocular interaction and depth sensitivity in striate and prestriate cortex of behaving rhesus monkey. J Neurophysiol. 1977;40:1392–1405. doi: 10.1152/jn.1977.40.6.1392. [DOI] [PubMed] [Google Scholar]
  32. Ponce CR, Lomber SG, Born RT. Integrating motion and depth via parallel pathways. Nat Neurosci. 2008 doi: 10.1038/nn2039. advance online publication, 13 Jan 2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Prince SJ, Pointon AD, Cumming BG, Parker AJ. Quantitative analysis of the responses of V1 neurons to horizontal disparity in dynamic random-dot stereograms. J Neurophysiol. 2002;87:191–208. doi: 10.1152/jn.00465.2000. [DOI] [PubMed] [Google Scholar]
  34. Qiu FT, von der Heydt R. Figure and ground in the visual cortex: v2 combines stereoscopic cues with gestalt rules. Neuron. 2005;47:155–166. doi: 10.1016/j.neuron.2005.05.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Ramsden BM, Hung CP, Roe AW. Real and illusory contour processing in area V1 of the primate: a cortical balancing act. Cereb Cortex. 2001;11:648–665. doi: 10.1093/cercor/11.7.648. [DOI] [PubMed] [Google Scholar]
  36. Roe A. Modular complexity of Area V2 in the macaque monkey. In: Kaas JH, Collins C, editors. The Primate Visual System. New York, NY: CRC Press; 2003. [Google Scholar]
  37. Roe AW, Fritsches K, Pettigrew JD. Optical imaging of functional organization of V1 and V2 in marmoset visual cortex. Anat Rec A Discov Mol Cell Evol Biol. 2005a;287:1213–1225. doi: 10.1002/ar.a.20248. [DOI] [PubMed] [Google Scholar]
  38. Roe AW, Lu HD, Hung CP. Cortical processing of a brightness illusion. Proc Natl Acad Sci U S A. 2005b;102:3869–3874. doi: 10.1073/pnas.0500097102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Roe AW, Ts'o DY. Visual topography in primate V2: multiple representation across functional stripes. J Neurosci. 1995;15:3689–3715. doi: 10.1523/JNEUROSCI.15-05-03689.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Sereno ME, Trinath T, Augath M, Logothetis NK. Three-dimensional shape representation in monkey cortex. Neuron. 2002;33:635–652. doi: 10.1016/s0896-6273(02)00598-6. [DOI] [PubMed] [Google Scholar]
  41. Shipp S, Zeki S. Segregation of pathways leading from area V2 to areas V4 and V5 of macaque monkey visual cortex. Nature. 1985;315:322–325. doi: 10.1038/315322a0. [DOI] [PubMed] [Google Scholar]
  42. Shipp S, Zeki S. The Organization of Connections between Areas V5 and V2 in Macaque Monkey Visual Cortex. Eur J Neurosci. 1989;1:333–354. doi: 10.1111/j.1460-9568.1989.tb00799.x. [DOI] [PubMed] [Google Scholar]
  43. Shmuel A, Korman M, Sterkin A, Harel M, Ullman S, Malach R, Grinvald A. Retinotopic axis specificity and selective clustering of feedback projections from V2 to V1 in the owl monkey. J Neurosci. 2005;25:2117–2131. doi: 10.1523/JNEUROSCI.4137-04.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Stepniewska I, Kaas JH. Topographic patterns of V2 cortical connections in macaque monkeys. J Comp Neurol. 1996;371:129–152. doi: 10.1002/(SICI)1096-9861(19960715)371:1<129::AID-CNE8>3.0.CO;2-5. [DOI] [PubMed] [Google Scholar]
  45. Swindale NV. How different feature spaces may be represented in cortical maps. Network. 2004;15:217–242. [PubMed] [Google Scholar]
  46. Thomas OM, Cumming BG, Parker AJ. A specialization for relative disparity in V2. Nat Neurosci. 2002;5:472–478. doi: 10.1038/nn837. [DOI] [PubMed] [Google Scholar]
  47. Ts'o DY, Roe AW, Gilbert CD. A hierarchy of the functional organization for color, form and disparity in primate visual area V2. Vision Res. 2001;41:1333–1349. doi: 10.1016/s0042-6989(01)00076-1. [DOI] [PubMed] [Google Scholar]
  48. Tsao DY, Vanduffel W, Sasaki Y, Fize D, Knutsen TA, Mandeville JB, Wald LL, Dale AM, Rosen BR, Van Essen DC, et al. Stereopsis activates V3A and caudal intraparietal areas in macaques and humans. Neuron. 2003;39:555–568. doi: 10.1016/s0896-6273(03)00459-8. [DOI] [PubMed] [Google Scholar]
  49. Wang Y, Xiao Y, Felleman DJ. V2 Thin Stripes Contain Spatially Organized Representations of Achromatic Luminance Change. Cereb Cortex. 2006 doi: 10.1093/cercor/bhj131. [DOI] [PubMed] [Google Scholar]
  50. Watanabe M, Tanaka H, Uka T, Fujita I. Disparity-selective neurons in area V4 of macaque monkeys. J Neurophysiol. 2002;87:1960–1973. doi: 10.1152/jn.00780.2000. [DOI] [PubMed] [Google Scholar]
  51. Xiao Y, Wang Y, Felleman DJ. A spatially organized representation of colour in macaque cortical area V2. Nature. 2003;421:535–539. doi: 10.1038/nature01372. [DOI] [PubMed] [Google Scholar]
  52. Xu X, Bosking W, Sary G, Stefansic J, Shima D, Casagrande V. Functional organization of visual cortex in the owl monkey. J Neurosci. 2004a;24:6237–6247. doi: 10.1523/JNEUROSCI.1144-04.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Xu X, Collins CE, Kaskan PM, Khaytin I, Kaas JH, Casagrande VA. Optical imaging of visually evoked responses in prosimian primates reveals conserved features of the middle temporal visual area. Proc Natl Acad Sci U S A. 2004b;101:2566–2571. doi: 10.1073/pnas.0308745101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Zeki S, Shipp S. Modular Connections between Areas V2 and V4 of Macaque Monkey Visual Cortex. Eur J Neurosci. 1989;1:494–506. doi: 10.1111/j.1460-9568.1989.tb00356.x. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS50688-supplement-01.pdf (820.7KB, pdf)

RESOURCES