Imagine you are looking out over a scenic landscape. The image you perceive is actually made up of many different visual components—for example color and movement—that are processed across many different areas in a region of the brain called the visual cortex. An important question for neuroscience is how the visual system combines information from so many different areas to create a coherent picture of the world around us.

Many areas of the visual cortex have their own map of what we see (the visual field). These maps allow the brain to maintain its representation of the visual field as the information passes from one processing area to the next. Areas that process corresponding parts of the visual field are physically interconnected, and tend to be active at the same time, which suggests that they are working together in some way. In addition, areas of the visual cortex that process different sections of the visual field can be activated at the same time, but it is not clear how this works.

Here, Arcaro et al. used a technique called functional magnetic resonance imaging (fMRI) to image the brains of people as they watched movies and while they rested. The images showed that seemingly unrelated areas of the visual cortex respond in similar ways if they are processing sections of the visual field that are the same distance from the center of the person's gaze. For example, if you look directly at the center of a computer screen parts of the brain that process the top of the screen are active at the same time as parts that process the bottom.

Arcaro et al.'s findings suggest that the brain uses the distance from the center of our gaze to bring together information from different areas of the visual cortex. This offers a new insight into how the brain assembles the many pieces of the visual jigsaw to make a complete picture. Future work will investigate how the architecture of the visual cortex is able to support this coupling of different areas, and how it might influence our perception of the visual world.

Within the visual system, a prominent organizational principle is that of retinotopy: adjacent neurons along the cortical surface typically receive input from adjacent points on the surface of the retina. Each retinotopic map can be divided along two orthogonal axes: polar angle (i.e., angular distance) and eccentricity (i.e., radial distance). Early electrophysiological recordings in monkeys and cats identified several representations of the contralateral visual field in and around the calcarine sulcus (Daniel and Whitteridge, 1961; Zeki, 1969; Allman and Kaas, 1971; Van Essen et al., 1984). Through the use of functional magnetic resonance imaging (fMRI), it has now become evident that the human visual system contains over two-dozen visual maps (for review of original mapping studies, see Silver and Kastner, 2009; Wandell and Winawer, 2011; Abdollahi et al., 2014; Wang et al., 2014).

The topographic organization of individual areas is thought to provide an infrastructure for the integration of information across areas and along the visual hierarchy (Kaas, 1997; Wandell et al., 2007). Anatomical studies in primates have demonstrated that neurons with overlapping receptive fields (RFs) are interconnected (Cragg, 1969; Zeki, 1969; Van Essen and Zeki, 1978; Maunsell and Van Essen, 1983). Similarly, fMRI connectivity studies in humans have demonstrated topographically-local correlations between regions with overlapping visual field representations (Heinzle et al., 2011; Haak et al., 2012; Butt et al., 2013; Donner et al., 2013; Gravel et al., 2014; Raemaekers et al., 2014). In addition, widespread functional correlation patterns have been observed across visual cortex in macaques (Leopold et al., 2003; Vincent et al., 2007) and humans (Nir et al., 2006, 2008; Yeo et al., 2011; Donner et al., 2013). These patterns contain broad differences between foveal and peripheral cortex (Raemaekers et al., 2014), though may also be tied to the fine-scale organization of individual retinotopic maps.

In this study, we used fMRI to investigate the relationship between the spatial pattern of correlated BOLD signal and the retinotopic organization of eight visual areas (V1–hV4, VO1–2, V3A–B). Correlation analyses were performed on data collected during task-free conditions (eyes-closed and fixation), and during movie viewing. Correlation patterns were consistent across subjects and experiments. In addition to finding patterns that support well-established anatomical connectivity between areas with overlapping RFs, our analyses revealed a widespread correlation pattern based on eccentricity representation, in which the BOLD signal was correlated in areas with non-overlapping visual field representations, but with matching eccentricity representations. This eccentricity-based correlation pattern was observed between upper and lower visual field representations, within and across visual areas, and between hemispheres. Moreover, correlation patterns were similar in the presence and absence of bottom-up visual input. Finally, the measured correlation pattern could not be accounted for by overlapping RFs, inter-hemisphere homotopic connections, anatomical distance, eye movements, subject motion, or physiological noise. Our results demonstrate that functional coupling between visual areas reflect both local and widespread topographical patterns. We propose that this widespread pattern is part of an intrinsic functional architecture of the visual system that could reflect eccentricity-dependent processing.

Retinotopy and correlation patterns were characterized and compared within the visual systems of 14 participants. Retinotopic organization of the visual system was examined within the central 15° of visual space using a conventional travelling wave paradigm in which eccentricity and polar angle maps were collected to define visual areas V1, V2, V3, V3A–B, hV4, VO1–2 (see ‘Materials and methods’). The organization of correlation patterns was probed in two resting conditions in which participants were instructed to either (1) keep eyes open and maintain fixation on a centrally presented dot or (2) keep eyes closed for the duration of the run. In addition, we assessed the correlation patterns in 11 of these participants during a naturalistic viewing condition in which participants were instructed to attend to a movie, but maintain fixation on a centrally presented dot.

Exploratory seed analysis

To investigate the topography of functional correlations across visual cortex, Pearson correlation coefficients were computed between the timeseries of all surface data points (nodes) within the left and right visual cortices. For any individual node, correlation coefficients with all other nodes within visual cortex typically ranged between −0.10 and 0.75 in individual subjects (after cerebrospinal fluid and white matter signal regression). For illustration of the raw functional correlation results, we present fixation resting-state correlation maps for four example seed locations in subject S1 (Figure 1A), eyes shut resting-state correlation maps for four example seed locations in subject S2 (Figure 1B), and group average resting-state maps for four example seed locations (Figure 2A). In each of the four seed locations, the BOLD timeseries was sampled from a single node (black dot) within right dorsal V2, and the seed location was gradually shifted from foveal (<1.0°) to peripheral-most (∼11.5°) representations as defined from a separate eccentricity localizer experiment (rightmost panel).

Figure 1 with 2 supplements see all

Download asset Open asset

Figure 2

Download asset Open asset

In all cases, we observed a combination of topographically local and widespread correlation patterns (with respect to visual field representations). Each of the four seed locations was strongly correlated (red / yellow) with adjacent cortex. In addition, the strongest correlations with each seed extended across visual cortex and spanned several visual areas, from V1 to V3A–B, dorsally, and to VO1–2 ventrally. Strong correlations with nodes adjacent to the seed and within ipsilateral dorsal cortex likely reflect well-established connectivity based on overlapping visual field representations (see Heinzle et al., 2011; Haak et al., 2012; Butt et al., 2013), but may also reflect the intrinsic spatial spread of the BOLD signal (Engel et al., 1997; Parkes et al., 2005). Interestingly, correlations were seen within both dorsal and ventral occipital cortex, comprised of lower and upper visual field representations respectively (Figure 1). Despite anatomical distance and representing a different part of visual space, the eccentricity representation (distance from fovea) of peak correlations within ventral occipital cortex corresponded to that of the seed location. These four seed locations also yielded comparable correlation patterns in the contralateral (left) hemisphere, comprised of right visual field representations. Local and widespread correlation patterns were observed in most individual subjects and the group average data for dorsal and ventral cortex seeds in V2 and V3, regardless of seeding near the horizontal or vertical meridians (see Figure 1—figure supplements 1, 2 for additional individual subject data).

The BOLD signals in areas with eccentricity preferences similar to the seed were correlated in the presence and absence of visual input, even in cases where the spatial receptive fields (RFs) were non-overlapping (i.e., across lower and upper or right and left visual field representations). To summarize the group average V2-seeds correlation results, we projected the correlation maps in Figure 2A into visual field coordinates, and averaged across areas V1, V2, V3, V3A–B, hV4, and VO1–2 (Figure 2B). The correlation patterns highlight both visuotopically local and widespread correlation patterns (Figure 2B). Peak correlations (red) were evident in parts of the visual field around each seed location with strong correlations (red / yellow) also extending across the visual field along an eccentricity ring corresponding to that of the seed location. Similar local and widespread eccentricity-based correlation patterns were observed in data from the movie viewing experiment (Figure 3). Individual subject and group average correlation patterns were similar to previously reported group average correlation patterns (Yeo et al., 2011). Below, we formally tested the relation of eccentricity representations to the spatial pattern of correlated BOLD signal between individual brain areas and across tasks.

Figure 3

Download asset Open asset

Eccentricity binning

To characterize the widespread eccentricity-based correlation pattern that was observed in raw correlation maps, individual subject timeseries data were grouped by visual area and then partitioned into 12 bins between 0.50° and 12.50° of eccentricity. Data binning was used as a form of averaging to increase signal-to-noise within bins, while preventing the spread of signal between bins. Within-subject pairwise correlations were calculated between the mean timeseries of all bins for visual areas V1, V2, V3, hV4, V3A–B (combined), and VO1–2 (combined), each of which has sufficient surface area to allow for fine-scale binning of eccentricity data. Areas V1–V3 were separated into quadrants for most analyses. No additional extrastriate areas were included in these analyses.

Correlations between dorsal and ventral bins of V1, V2, and V3 were strongest for iso-eccentricity representations. These correlations were seen across the vertical and horizontal meridians at both foveal and peripheral-most bins. We first illustrate the binning analysis by examining the raw correlations between the ventral and dorsal quadrants in V3 (Figure 4). This pair was chosen for illustration because correlations computed across quadrants are less susceptible to the influence of overlapping receptive fields (RFs) or cortical proximity. For regions that represent the same visual quadrant (e.g., the dorsal portions of visual areas V1, V2, and V3), it is difficult to dissociate effects due to shared eccentricity representations from overlapping receptive fields. Similarly, for adjacent dorsal regions, the shortest cortical distances are typically at corresponding eccentricity representations, making it difficult to dissociate eccentricity-related correlations from cortical (and volumetric) distance-based correlations. The dorsal and ventral portions of visual area V3 (as well as V2), however, only anatomically border each other at the fovea, represent different parts of the visual field (lower and upper, respectively), and thereby minimize both the overlapping receptive field and anatomical adjacency concerns. Thus, correlation analyses between these areas allowed us to test for widespread eccentricity-based correlation patterns in cases where effects of cortical distance and overlapping RFs are minimal. As seen for subject S4, the exploratory seed-based analyses showed strong correlations (red/yellow) between dorsal and ventral V3 at corresponding eccentricity representations (Figure 4A). Binned data showed the same correlation pattern. For each of the selected dorsal V3 bins, correlations with ventral V3 bins were strongest at (and around) corresponding iso-eccentricities (Figure 4B). For example, peripheral-most V3d bins correlated most with peripheral-most V3v bins, and the correlations gradually decreased moving towards the foveal bin (Figure 4B, right). The entire correlation matrix for all eccentricity locations between dorsal and ventral V3 (Figure 4C, center panel), revealed a similar pattern, where correlations were strongest among ventral and dorsal bins with iso-eccentricity representations (i.e., the diagonal), and were weaker for bins with large radial distances (e.g., foveal vs peripheral-most).

Figure 4

Download asset Open asset

The pattern of correlations reflects radial distance, not anatomical distance. We computed the (ranked) radial distance between eccentricity bins. A radial bin distance of zero corresponds to bins with the same (iso-) eccentricity representation, while a radial bin distance of 11 corresponds to bins furthest from each other on the eccentricity axis (i.e., foveal vs peripheral-most). Figure 4C provides the spatial pattern of correlations between ventral and dorsal quadrants in area V3 as well as the predicted patterns based on radial distance (left gray panel) and based on cortical distance (right gray panel). Note that both the radial bin distance and the cortical distance (as well as an overlapping RF model) predict strong correlations between foveal bins (Figure 4C). However, whereas cortical distance predicts that correlations will be weaker between cortically–distant iso-eccentricity bins (e.g., ventral and dorsal V3 peripheral-most), the radial bin distance predicts a strong correlation between iso-eccentricities (Figure 4C). Indeed, the group average bin data was strongly correlated with the predicted eccentricity pattern (r = 0.84) and was not positively correlated with cortical distance (r = −0.04) or volume distance (r = −0.19) (Figure 4C). Individual subject matrices were also correlated with radial distance (mean r = 0.45), and the Fisher-transformed correlations significantly differed from zero across subjects (one-sample t-test, t(13) = 6.37, p < 0.0001).

Correlations with radial bin distance were apparent within and across all visual areas tested, including areas with overlapping and non-overlapping visual field representations (both within and across hemispheres). We computed intra- and inter-hemisphere correlations as a function of eccentricity bin within and across visual areas V1, V2, V3, hV4, V3A–B (combined), and VO1–2 (combined) in each condition (resting eyes closed, resting eyes open, and movie fixation; Figure 5). In Figure 5, gray frames denote comparisons between regions with overlapping visual field representations while black frames denote comparisons between regions with minimal or no overlap in visual field representations (i.e., across ventral and dorsal visual fields and across hemispheres). For all comparisons in each condition the strongest correlations (red/yellow) were consistently between bins at comparable eccentricity locations (i.e., along the diagonal of each sub-matrix in Figure 5). This was the case even for correlations between dorsal and ventral visual portions of V1, V2, and V3, and between hemispheres, which represent mostly discrete parts of visual space. Across area pairs, the average individual subject correlation coefficient with radial distance ranged between 0.17 and 0.80 with a median of 0.39 and an interquartile range of 0.12. Fisher-transformed individual subject correlations significantly differed from zero for all pairs in each experiment (one-sample t-test, t(13) > 2.31, ps < 0.05, FDR-corrected) except for two pairs of inter-hemisphere correlations with hV4 in the movie viewing condition (ps = 0.13 and 0.07). In all experiments, individual subject correlations with radial distance were generally strongest for early visual areas (V1, V2, and V3). This could reflect the relatively smaller, more spatially focal receptive fields in early visual areas, but also the larger surface area (i.e., more data points).

Figure 5

Download asset Open asset

Effect of radial distance

For each area pair, correlation coefficients were positive at 0–radial distance, and decreased at larger radial distances across the visual field. To assess the general effect of radial distance for each area pair, eccentricity bin correlations were grouped as a function of radial bin distance and averaged. Averaged correlations were plotted relative to iso-eccentricity (Figure 6A, left panel). In each quadrant, the radial mid-arc (Figure 6A, white outline) corresponds to the average correlation of the 0-distance bin (iso-eccentricity). Radial arcs further away from the mid-arc correspond to larger radial distances. For each radial plot, for example, V2–V3, radial arcs inside of the mid-arc illustrate correlations between relatively more foveal representations of V2 and relatively more peripheral representations of V3. Radial arcs outside of the mid-arc illustrate correlations between relatively more peripheral representations of V2 and relatively more foveal representations of V3. Inner and outer radial points will be identical for within area comparisons (e.g., V2–V2), but could differ for between area comparisons (e.g., V2–V3). As illustrated by resting state correlations within and between areas V2 and V3 (Figure 6A), iso-eccentricity bins were positively correlated, regardless of visual field quadrant. This was observed in all experiment conditions and area comparisons, and individual subject Fisher-transformed correlations significantly differed from zero (ps < 0.05, FDR corrected). There were clear differences in overall correlation magnitude between quadrants. As expected, correlation coefficients decreased at larger radial distances for foveal and peripheral-most bins in each quadrant, regardless of this magnitude difference (Figure 6A).

Figure 6 with 3 supplements see all

Download asset Open asset

The relative decrease in correlation strength as a function of radial distance was similar across the visual field and across experiments (Figure 6B). We plotted the average correlation as a function of radial bin distance (relative to iso-eccentricity; see ‘Materials and methods’) between the upper and lower visual fields, right and left visual fields, as well as across both left and right and upper and lower visual fields separately (see graphic illustrations in Figure 6B left panels). Data are presented for comparisons between quadrants within and between areas V2 and V3. As discussed in the eccentricity binning section, these comparisons were chosen for illustration because overlap in visual field representation is minimal. Consistent with the radial arc plots (Figure 6A), correlations between dorsal and ventral portions of V2 and V3 were strongest at 0–radial distance, and steadily decreased at larger radial distances for all conditions (Figure 6; top row). We observed similar correlation patterns as a function of radial bin distance across hemispheres, both along the horizontal plane (e.g., RH V2v and LH V2v; middle row) and diagonally across the upper and lower visual fields (e.g., RH V2v and LH V2d; bottom row). The patterns were observed even after removal of horizontal and vertical meridian data, further demonstrating that these effects are unlikely to be driven by local, overlapping visual field representations (Box 1; Figure 6—figure supplement 1). The decrease in correlation coefficients at larger radial bin distances yielded negative slopes from linear fits in each subject for all V2 and V3 pairs. For any given pair, negative slopes were similar across condition. Across V2 and V3 pairs, the average individual subject slope of the linear fit ranged between −0.22 and −0.49. Further, the average individual subject slope for each area pair was greater than 97.5% of a permuted distribution (mean 97.5% across areas, conditions = −0.06) where the labels of radial distances were scrambled before deriving individual subject radial distance correlations and slopes. As seen in the radial plots (Figure 6B), the slopes were similar across regional comparisons, though were slightly shallower for comparisons that spanned both quadrants and hemispheres (i.e., diagonal; bottom row). Finally, we observed these effects during resting fixation and eyes closed conditions, as well as during the processing of the movie, attesting to the robustness of the effect, and excluding many potential confounds (see ‘Control analyses’). Similar results were found in all other comparisons between visual areas. The magnitude of correlation coefficients linearly decreased as a function of radial bin distance for all area comparisons (V1, V2, V3, hV4, VO1–2, V3A–B), though slopes were generally steeper for comparisons between V1, V2, and V3. Across all areas, average individual subject slopes ranged between −0.14 and −0.80, and each was greater than 97.5% of the permuted distribution (mean 97.5% across areas, conditions = −0.09).

Box 1

Widespread connectivity is not dependent on meridian representations.

Though we divided data into quadrants that represent distinct parts of visual space, monosynaptic connections have been observed between neurons in dorsal and ventral visual cortex with receptive fields overlapping at the horizontal meridian (Jeffs et al., 2009; see also; Zeki, 1971; Stepniewska and Kaas, 1996; Felleman et al., 1997; Gattass et al., 1997) and between hemispheres at the vertical meridian (Essen and Zeki, 1978; Newsome and Allman, 1980; Cusick et al., 1984; Kennedy et al., 1986; Abel et al., 2000). Despite constituting (at most) a minimal portion of the signal in our bin data, we tested whether the observed eccentricity-based correlations between dorsal and ventral cortex and between the hemispheres were driven by the inclusion of these overlapping representations at the horizontal and vertical meridians, respectively, by removing data from the meridians and re-running the binning analyses. For correlations between upper and lower visual fields (dorsal and ventral cortex), we cut out 60° of polar angle centered on the horizontal meridian, which spared 60° of polar angle in the upper and lower visual fields near the vertical meridians. For correlations between the right and left visual fields (right and left hemispheres), we cut out 90° of polar angle centered on each vertical meridian, which sparred 90° of polar angle centered on the horizontal meridian in each hemisphere. Consistent with the original analyses, average individual subject correlations between dorsal and ventral portions of V2 and V3 as well as between right and left hemispheres were strongest at the 0–radial distances (iso-eccentricities), and linearly decreased at larger distances for all conditions (Figure 6—figure supplement 1). In general, the slopes of correlations were marginally shallower, but did not significantly differ from the full data reported in Figure 5 (ps > 0.05, FDR corrected). If overlapping meridian representations facilitated the correlation patterns, removal of this data should have eliminated the correlation effects. These data further support the interpretation that the observed correlations with radial distance are not due to overlapping representations of visual space.

https://doi.org/10.7554/eLife.03952.014

Stimulus-dependent and independent correlations

For each area pair, the spatial pattern of correlations (across eccentricity bins) was similar between experimental conditions. Across all areas, the average individual subject correlation of these spatial patterns between experimental conditions was 0.86 for intra-hemisphere comparisons and 0.75 for inter-hemisphere comparisons. The average individual subject correlation between resting-state and movie viewing conditions was 0.84 for intra-hemisphere comparisons and 0.71 for inter-hemisphere comparisons. Overall, these data are consistent with the exploratory seed based analyses, and suggest that the spatial pattern of correlations between visual areas tested were similar in the presence and absence of a strong bottom-up input.

Given the similarity of correlation matrices across all three conditions, this widespread eccentricity-based correlation pattern appears to reflect an eccentricity bias that is inherent to the organization of the visual system (i.e., stable during rest) and may support processing during active perception of the visual environment (e.g., during movie viewing). We conducted correlations between individual runs to directly test whether the patterns observed during the movie viewing condition reflected processing of the stimulus input. Intrinsic neural dynamics during the resting and movie conditions that are not related to the processing of visual stimuli, as well as non-neuronal artifacts (e.g., respiratory rate, motion), can only influence the pattern of correlations within each run, but should not induce correlations between runs. Indeed, inter-run Fisher-transformed correlations on resting state data showed no effect of radial distance (all ps > 0.05), validating the assumption that noise correlations should not be reliable across runs (Figure 7, blue and black lines). In contrast, inter-run correlations during movie watching showed radial distance effects similar to intra-run correlations (Figure 7, red lines). Inter-run correlations were statistically significant for 96% of all comparisons (ps < 0.05, FDR corrected). The slopes of coefficients as a function of radial distance for movie data were, generally, weaker for the inter-run relative to intra-run analyses (compare Figures 6B, 7). However, the inter-run analyses inherently had less data and spatial attention was not controlled, making the differences between intra- and inter-run analyses difficult to interpret. The inter-run results indicate that the observed widespread eccentricity-based correlations can reflect the processing of the incoming information during viewing of real life stimuli, and further suggest that these correlations patterns are unlikely to be driven by non-neuronal artifacts.

Figure 7

Download asset Open asset

Topographic model regression analyses

Both local and widespread connectivity patterns contributed to the spatial pattern of observed correlations. Each visual hemifield map was separated into 36 bins: six divisions of eccentricity (E1–E6) that each contained six divisions of polar angle (A1–A6) such that each bin represented a unique part of visual space (Figure 8—figure supplement 1). To visualize the pattern of correlations across all bins with respect to local (overlapping RF) and widespread (eccentricity and polar angle) connectivity patterns, data were plotted as a function of radial and angular distance (Figure 8). Correlations were qualitatively strongest at the intersection of iso-polar angle and iso-eccentricity, and were generally elevated along iso-eccentricity representations, irrespective of polar angle distance (Figure 8). To quantify this, we assessed the relationship between the measured correlations across visual areas V1, V2, V3, and hV4 and several possible connectivity patterns using linear, least squares regression (see ‘Materials and methods’). For any pair of visual areas, we modeled correlations between all bin pairs (36 × 36) as the linear weighted sum of four sources of connectivity. Two sources reflect topographically local connectivity between regions that are in close anatomical proximity or contain overlapping receptive fields: (1) instrumental ‘noise’ connectivity: correlation pattern based on an assessment of the spatial auto-correlation and preprocessing in our data; (2) overlapping RF connectivity: correlation pattern based on the overlap of estimated population receptive fields (pRF; Dumoulin and Wandell, 2008; Amano et al., 2009; Harvey and Dumoulin, 2011). The two other sources reflect topographically widespread connectivity that span much of the visual field and are not specifically tied to receptive fields or anatomical proximity: (1) eccentricity connectivity: correlation pattern based on radial distance with correlations strongest at iso-eccentricity representations; (2) polar angle connectivity: correlation pattern based on angular distance with correlations strongest at iso-polar angle representations.

Figure 8 with 2 supplements see all

Download asset Open asset

In general, the strongest overlapping RF effects were observed for intra-areal comparisons whereas eccentricity effects were consistently observed across all comparisons and were much larger than overlapping RF and polar angle effects for inter-hemisphere comparisons. To quantify the apparent connectivity patterns, we conducted an initial regression analysis using a model that included all four sources of connectivity as predictors. For individual subjects, the combination of local and widespread predictors well fit intra-area correlation patterns, and moderately fit inter-area and inter-hemisphere patterns. The average individual subject variance explained by the full model ranged between 40% and 68% for intra-areal comparisons and between 11% and 28% for inter-areal comparisons. Across area pairs, there were consistent effects of both local and widespread predictors (Figure 9—figure supplement 1). Across subjects, the coefficients for the overlapping RF predictor were significantly positive for all intra-hemisphere, intra-areal comparisons as well as for most other comparisons (ps < 0.05, FDR-corrected). Coefficients for the eccentricity predictor were significantly positive for all but one comparison (V1–hV4 inter-hemisphere) (ps < 0.05, FDR corrected). Coefficients for the polar angle predictor were significantly positive for most intra-hemisphere comparisons, but only for a few inter-hemisphere comparisons (ps < 0.05, FDR corrected). Though these results suggest some effect of both local and widespread connectivity, we found that both eccentricity and polar angle predictors were strongly correlated with the overlapping RF predictor for most intra-hemisphere area pairs (mean r = 0.54; STD = 0.08). Since the local and widespread predictors have a non-trivial degree of collinearity, the coefficient estimates for individual predictors from the full model may not be accurate for intra-hemisphere comparisons, which had the strongest correlations between local and widespread predictors. From these results, it is possible that some of the observed eccentricity and polar angle effects are due to shared variance with the local, overlapping RF model. However, it is very unlikely that all of the widespread effects can generically be explained by the overlapping receptive field model since significant correlations were observed in the eccentricity bin analyses between iso-eccentricity representations in distinct parts of visual space (Figure 6; e.g., between V2d RH and V2v LH).

When removing the shared variance between local and widespread connectivity predictors, significant effects of eccentricity connectivity were still observed. To assess the contribution of widespread connectivity on the measured correlation patterns while controlling for the shared variance with the local connectivity model, we first removed the variance explained by local connectivity (i.e., overlapping RF and instrumental predictors), then calculated correlations between the residuals and the widespread connectivity models in each subject. The average correlation coefficient between the residual pattern and the eccentricity predictor was similar across most areas for both within and across hemispheres (Figure 9, medium gray bars). The Fisher-transformed individual subject eccentricity residual correlations were reliably different from zero for 56/60 visual area pairs across rest and movie viewing experiments (ps < 0.05, FDR corrected). Consistent with the binning analyses, residual correlations that included hV4 were weaker than residual correlations between areas V1, V2, and V3. In contrast, there was no clear, consistent effect of the polar angle predictor across areas (Figure 9, lightest gray bars). The polar angle residual correlations were only significantly positive for 4/60 pairs (ps < 0.05, FDR corrected). For all other comparisons, residual polar angle correlations tended to be negative. The residual pattern was more correlated with the eccentricity predictor than with the polar angle predictor for 49/60 pairs (ps < 0.05, FDR corrected) (Figure 9).

Figure 9 with 3 supplements see all

Download asset Open asset

These residual widespread effects were robust to scaling of pRF sizes in the local connectivity model. We used the average pRF values reported across several studies to derive our model of overlapping RF connectivity, however the size of pRFs for some areas varied as much as 2× across these studies. To make sure that the variance in pRF estimates across studies didn't significantly affect the results, we re-ran the model analyses using pRF estimates from individual studies to generate the model of overlapping RFs (see ‘Materials and methods’). Using these parameters to derive the overlapping RF model had no appreciable effect on the magnitude or significance of residual correlations. Further, the magnitude and significance of the eccentricity-based effects were also largely unchanged when we scaled pRF estimates to account for any additional effects of the point spread function in the BOLD signal not captured in the original pRF estimate (see ‘Materials and methods’; Figure 9—figure supplement 2). Overall, these control analyses provide additional evidence for a widespread eccentricity-based connectivity that cannot be accounted for by local connectivity patterns.

We investigated the spatial pattern of functional correlations across eight human early visual and extrastriate cortical areas using fMRI during conditions in which there was little or no external input (resting-state) and during conditions in which there was a dynamic external input (movie viewing). Local and widespread (spatial) patterns of correlated BOLD signal were observed in all experiments, both in the presence and absence of visual input. In agreement with prior reports, we observed topographically-local correlation patterns based on overlapping representations of visual space. In addition, we found strong evidence for topographically-widespread correlation patterns based on eccentricity within and across hemispheres, which spanned the entire visual field. The eccentricity-based correlation patterns were strongest between early visual areas (V1–V3). The effects observed in extrastriate areas (hV4, V3A–B, and VO1–2) were generally weaker than early visual areas, consistent with the coarser topographic organization of visual space and smaller surface area in these areas.

Our data extend recent findings of topographic connectivity within visual cortex using fMRI. At a fine-scale, correlation patterns reflect overlapping RFs within hemispheres and homotopic connections between hemispheres (Heinzle et al., 2011; Haak et al., 2012; Butt et al., 2013). By modeling V1 inter-areal connections based on overlapping RFs, Gravel and colleagues (2014) generated polar angle and eccentricity maps for V2 and V3. However, Raemaekers and colleagues (2014) reported that such fine-scaled connectivity was only observable after filtering coarse-scale components. At a coarse scale, there has been evidence for a general foveal-peripheral distinction (Vincent et al., 2007; Nallasamy and Tsao, 2011; Smith et al., 2012; Raemaekers et al., 2014), but these studies did not report a systematic topography of correlation patterns that reflects the underlying eccentricity organization, nor were effects of overlapping RFs and cortical distance controlled. Evidence for a more systematic relation to eccentricity was reported between V1 and ventral V3 (Yeo et al., 2011), though such data are also consistent with overlapping RF connectivity as well as cortical distance since only ventral V3 was probed. Here, we explicitly and quantitatively combined these connectivity phenomena in a fine-grained analysis across a wider range of brain regions, replicating previous findings of overlapping RF and homotopic correlation patterns without filtering coarse spatial components (Heinzle et al., 2011; Haak et al., 2012; Butt et al., 2013; Gravel et al., 2014). Even after accounting for these patterns, we found a robust widespread correlation pattern that reflects the eccentricity organization across much of visual cortex. Such systematic correlation patterns suggest orderly integration processes across the whole visual field at multiple levels of the processing hierarchy, not just a foveal-peripheral dichotomy.

Eccentricity-based correlation patterns may reflect an intrinsic functional organization of visual cortex. Our results during rest demonstrate that regions with iso-eccentricity representations are likely to be co-active, even in the absence of visual input. Our results during the movie viewing condition demonstrate that the temporal dynamics yielding these eccentricity-based correlation patterns are also present during strong bottom-up input, and indicate that this organization is relevant for the processing of incoming visual input. In agreement with these findings, studies in macaques and cats have shown that the activity of neurons with similar response properties are correlated in the presence and absence of external input, suggesting that spontaneous neuronal activity is tightly linked to intrinsic cortical networks (Arieli et al., 1995; Tsodyks et al., 1999; Goldberg et al., 2004; Ghuman et al., 2013). Further, both overlapping RF and eccentricity-based connectivity patterns were observed in the presence and absence of external input, suggesting that topographically-local and widespread patterns are both part of this intrinsic functional architecture.

The observed patterns of functional connectivity may reflect both direct and indirect anatomical connectivity (Vincent et al., 2007; Honey et al., 2009). Local connectivity is likely supported by direct anatomical connections between overlapping RFs (Cragg, 1969; Essen and Zeki, 1978; Maunsell and Van Essen, 1983). Such wiring is necessary for the integration of information within focal points of our visual environment. Direct intra-areal anatomical connections between dorsal and ventral visual cortex at the horizontal meridians (Jeffs et al., 2009) and between both hemispheres at the vertical meridians (Hubel and Wiesel, 1967; Essen and Zeki, 1978; Newsome and Allman, 1980; Cusick et al., 1984; Kennedy et al., 1986) could support widespread functional connectivity across the visual field, though we are not aware of anatomical studies explicitly reporting eccentricity-based patterns of intra-areal connectivity. Further, while labeled cells of lateral connections in macaque striate and extrastriate cortex exhibit some anisotropy with respect to the cortical surface, this is thought to reflect cortical magnification factor, and yield isotropic visual field coverage (Angelucci et al., 2002). Consistent with the anatomical connectivity, we found that correlation patterns between regions with overlapping RFs were comparable along eccentricity and polar angle dimensions. Beyond overlapping RFs, correlation patterns were anisotropic (with respect to visual field coverage) and reflected the underlying eccentricity organization. Alternatively, the observed eccentricity-based correlation patterns may actually reflect a broader-scale anatomical organization of direct (and indirect) connections, facilitated via differences in intra-areal projections between cortical sites representing central and peripheral space (Colby et al., 1988; Nakamura et al., 1993; Gattass et al., 2005; Ungerleider et al., 2008, 2014). Such a distinction has been characterized in the patterns of supra-areal anatomical connections between early visual and extrastriate cortex in non-human primates (Rosa, 2002; Gattass et al., 2005; Rosa and Tweedale, 2005; Rosa et al., 2009; Buckner and Yeo, 2014). It is not known whether these anatomical connectivity patterns are ‘bi-modal’, and only distinguish central and peripheral space, or reflect a finer-scale organization where connectivity patterns with intermediate eccentricity representations are distinguishable from central and peripheral connectivity profiles. Our results predict that these anatomical connections should reflect a gradient, though this remains to be explored. In particular, feedback projections from extrastriate areas with receptive fields covering wide swaths of the visual field to early and intermediate visual areas could facilitate such widespread, eccentricity-dependent correlation patterns. It is interesting to note that when comparing the profile of anatomical connectivity between V2/V4 and higher order cortex (e.g., Figure 7, Gattass et al., 2005) with the organization of eccentricity across visual cortex in macaques (Brewer et al., 2002; Kolster et al., 2009; Arcaro et al., 2011), it is clear that higher order areas connected with peripheral parts of V2 and V4 (e.g., PO, PIP, LIP, DP, TF) have a large representation of the periphery, and higher order areas connected with foveal parts of V2 and V4 (e.g., TEO and TE) have a large representation of the fovea. The exact relationship between the observed correlation patterns and anatomical pathways will need to be further investigated.

Our data link the functional organization of early and higher order visual cortex. Previous studies have proposed eccentricity as a large-scale functional organizing principle for higher order visual cortex (Levy et al., 2001; Hasson et al., 2002; Malach et al., 2002). Higher order areas with foveal biases tend to be specialized in face and object recognition, and areas with peripheral biases tend to be involved in scene analysis (Levy et al., 2001; Hasson et al., 2002; Malach et al., 2002). Perceptually, these recognition processes require different visual acuities. For example, while fine acuity is needed for the featural discrimination among similar face and object exemplars (Fiorentini et al., 1983; Goffaux et al., 2005; Keil, 2008), a coarser acuity is needed for mapping the surrounding layout necessary for navigation in space (Oliva and Schyns, 1997; Oliva and Torralba, 2006). Further, eye movement patterns during scene perception are related to the types of information within a scene (Buswell, 1935; Henderson and Hollingworth, 1999). People tend to foveate on faces while orienting their peripheral vision at landscape features and room contours (Yarbus, 1967). These eccentricity biases are also reflected in the connectivity patterns between face and place category-selective regions in ventral temporal cortex with extrastriate visual area hV4 (Baldassano et al., 2012). Our data show that this divergence in the computational processes necessary for foveal and peripheral recognition is evident even in early visual cortex.

Our results underscore the importance of relating functional connectivity data to known functional (or anatomical) organization (Jbabdi et al., 2013; Sporns and Honey, 2013; Wang et al., 2013). The detailed retinotopic organization of visual cortex allowed for a unique opportunity to systematically compare patterns of correlated BOLD activity with the known underlying functional organization of the visual system. Across subjects and experiments, correlations were stronger between areas at matched eccentricities than within areas at large eccentricity distances, suggesting that functional connectivity analyses on BOLD data are more sensitive at revealing widespread, inter-areal connectivity patterns than the localization of individual retinotopic areas (see also Yeo et al., 2011). Alternative connectivity approaches not based on similarity may prove useful at revealing area boundaries (Wig et al., 2014), though this remains to be tested more thoroughly beyond the V1/V2 border. Thus, we propose that relating correlation patterns to known functional and anatomical data will prove important for identifying the functional pathways for the integration of information across individual, functionally specialized areas.

1. 1. Abdollahi RO
   2. Kolster H
   3. Glasser MF
   4. Robinson EC
   5. Coalson TS
   6. Dierker D
   7. Jenkinson M
   8. Van Essen DC
   9. Orban GA

   (2014) Correspondences between retinotopic areas and myelin maps in human visual cortex

   Neuroimage 99:509–524.

   https://doi.org/10.1016/j.neuroimage.2014.06.042

   • Google Scholar
2. 1. Abel PL
   2. O'brien BJ
   3. Olavarria JF

   (2000) Organization of callosal linkages in visual area V2 of macaque monkey

   The Journal of Comparative Neurology 428:278–293.

   https://doi.org/10.1002/1096-9861(20001211)428:2<278::AID-CNE7>3.0.CO;2-R

   • Google Scholar
3. 1. Allman JM
   2. Kaas JH

   (1971) Representation of the visual field in striate and adjoining cortex of the owl monkey (Aotus trivirgatus)

   Brain Research 35:89–106.

   https://doi.org/10.1016/0006-8993(71)90596-8

   • Google Scholar
4. 1. Amano K
   2. Wandell BA
   3. Dumoulin SO

   (2009) Visual field maps, population receptive field sizes, and visual field coverage in the human MT+ complex

   Journal of Neurophysiology 102:2704–2718.

   https://doi.org/10.1152/jn.00102.2009

   • Google Scholar
5. 1. Angelucci A
   2. Levitt JB
   3. Walton EJ
   4. Hupe JM
   5. Bullier J
   6. Lund JS

   (2002)

   Circuits for local and global signal integration in primary visual cortex

   The Journal of Neuroscience 22:8633–8646.

   • Google Scholar
6. 1. Arcaro MJ
   2. Mcmains SA
   3. Singer BD
   4. Kastner S

   (2009) Retinotopic organization of human ventral visual cortex

   The Journal of Neuroscience 29:10638–10652.

   https://doi.org/10.1523/JNEUROSCI.2807-09.2009

   • Google Scholar
7. 1. Arcaro MJ
   2. Pinsk MA
   3. Li X
   4. Kastner S

   (2011) Visuotopic organization of macaque posterior parietal cortex: a functional magnetic resonance imaging study

   The Journal of Neuroscience 31:2064–2078.

   https://doi.org/10.1523/JNEUROSCI.3334-10.2011

   • Google Scholar
8. 1. Argall BD
   2. Saad ZS
   3. Beauchamp MS

   (2006) Simplified intersubject averaging on the cortical surface using SUMA

   Human Brain Mapping 27:14–27.

   https://doi.org/10.1002/hbm.20158

   • Google Scholar
9. 1. Arieli A
   2. Shoham D
   3. Hildesheim R
   4. Grinvald A

   (1995)

   Coherent spatiotemporal patterns of ongoing activity revealed by real-time optical imaging coupled with single-unit recording in the cat visual cortex

   Journal of Neurophysiology 73:2072–2093.

   • Google Scholar
10. 1. Baldassano C
    2. Iordan MC
    3. Beck DM
    4. Fei-Fei L

    (2012) Voxel-level functional connectivity using spatial regularization

    Neuroimage 63:1099–1106.

    https://doi.org/10.1016/j.neuroimage.2012.07.046

    • Google Scholar
11. 1. Bandettini P

    (2009)

    Neural correlates of thinking

    Neural correlates of thinking, Berlin, Heidelberg, Springer.

    • Google Scholar
12. 1. Benjamini Y
    2. Hochberg Y

    (1995)

    Controlling the false discovery rate: a practical and powerful approach to multiple testing

    Journal of Royal Statistical Society, Series B 57:289–300.

    • Google Scholar
13. 1. Biswal B
    2. Yetkin FZ
    3. Haughton VM
    4. Hyde JS

    (1995) Functional connectivity in the motor cortex of resting human brain using echo-planar MRI

    Magnetic Resonance in Medicine 34:537–541.

    https://doi.org/10.1002/mrm.1910340409

    • Google Scholar
14. 1. Brewer AA
    2. Liu J
    3. Wade AR
    4. Wandell BA

    (2005) Visual field maps and stimulus selectivity in human ventral occipital cortex

    Nature Neuroscience 8:1102–1109.

    https://doi.org/10.1038/nn1507

    • Google Scholar
15. 1. Brewer AA
    2. Press WA
    3. Logothetis NK
    4. Wandell BA

    (2002)

    Visual areas in macaque cortex measured using functional magnetic resonance imaging

    The Journal of Neuroscience 22:10416–10426.

    • Google Scholar
16. 1. Buckner RL
    2. Yeo BT

    (2014) Borders, map clusters, and supra-areal organization in visual cortex

    Neuroimage 93:292–297.

    https://doi.org/10.1016/j.neuroimage.2013.12.036

    • Google Scholar
17. Book

    1. Buswell GT

    (1935)

    How people look at pictures: a study of the psychology and perception in art

    Oxford, England: Chicago Press.

    • Google Scholar
18. 1. Butt OH
    2. Benson NC
    3. Datta R
    4. Aguirre GK

    (2013) The fine-scale functional correlation of striate cortex in sighted and blind people

    The Journal of Neuroscience 33:16209–16219.

    https://doi.org/10.1523/JNEUROSCI.0363-13.2013

    • Google Scholar
19. 1. Chung MK
    2. Robbins SM
    3. Dalton KM
    4. Davidson RJ
    5. Alexander AL
    6. Evans AC

    (2005) Cortical thickness analysis in autism with heat kernel smoothing

    Neuroimage 25:1256–1265.

    https://doi.org/10.1016/j.neuroimage.2004.12.052

    • Google Scholar
20. 1. Colby CL
    2. Gattass R
    3. Olson CR
    4. Gross CG

    (1988) Topographical organization of cortical afferents to extrastriate visual area PO in the macaque: a dual tracer study

    The Journal of Comparative Neurology 269:392–413.

    https://doi.org/10.1002/cne.902690307

    • Google Scholar
21. 1. Cox RW

    (1996) AFNI: software for analysis and visualization of functional magnetic resonance neuroimages

    Computers and Biomedical Research 29:162–173.

    https://doi.org/10.1006/cbmr.1996.0014

    • Google Scholar
22. 1. Cragg BG

    (1969) The topography of the afferent projections in the circumstriate visual cortex of the monkey studied by the Nauta method

    Vision Research 9:733–747.

    https://doi.org/10.1016/0042-6989(69)90011-X

    • Google Scholar
23. 1. Cusick CG
    2. Gould HJ III
    3. Kaas JH

    (1984) Interhemispheric connections of visual cortex of owl monkeys (Aotus trivirgatus), marmosets (Callithrix jacchus), and galagos (Galago crassicaudatus)

    The Journal of Comparative Neurology 230:311–336.

    https://doi.org/10.1002/cne.902300302

    • Google Scholar
24. 1. Dale AM
    2. Fischl B
    3. Sereno MI

    (1999) Cortical surface-based analysis. I. Segmentation and surface reconstruction

    Neuroimage 9:179–194.

    https://doi.org/10.1006/nimg.1998.0395

    • Google Scholar
25. 1. Daniel PM
    2. Whitteridge D

    (1961) The representation of the visual field on the cerebral cortex in monkeys

    The Journal of Physiology 159:203–221.

    https://doi.org/10.1113/jphysiol.1961.sp006803

    • Google Scholar
26. 1. DeYoe EA
    2. Carman GJ
    3. Bandettini P
    4. Glickman S
    5. Wieser J
    6. Cox R
    7. Miller D
    8. Neitz J

    (1996) Mapping striate and extrastriate visual areas in human cerebral cortex

    Proceedings of the National Academy of Sciences of USA 93:2382–2386.

    https://doi.org/10.1073/pnas.93.6.2382

    • Google Scholar
27. 1. Donner TH
    2. Sagi D
    3. Bonneh YS
    4. Heeger DJ

    (2013) Retinotopic patterns of correlated fluctuations in visual cortex reflect the dynamics of spontaneous perceptual suppression

    The Journal of Neuroscience 33:2188–2198.

    https://doi.org/10.1523/JNEUROSCI.3388-12.2013

    • Google Scholar
28. 1. Dumoulin SO
    2. Wandell BA

    (2008) Population receptive field estimates in human visual cortex

    Neuroimage 39:647–660.

    https://doi.org/10.1016/j.neuroimage.2007.09.034

    • Google Scholar
29. 1. Engel SA
    2. Glover GH
    3. Wandell BA

    (1997) Retinotopic organization in human visual cortex and the spatial precision of functional MRI

    Cerebral Cortex 7:181–192.

    https://doi.org/10.1093/cercor/7.2.181

    • Google Scholar
30. 1. Felleman DJ
    2. Burkhalter A
    3. Van Essen DC

    (1997) Cortical connections of areas V3 and VP of macaque monkey extrastriate visual cortex

    The Journal of Comparative Neurology 379:21–47.

    https://doi.org/10.1002/(SICI)1096-9861(19970303)379:13.0.CO;2-K

    • Google Scholar
31. 1. Fiorentini A
    2. Maffei L
    3. Sandini G

    (1983) The role of high spatial frequencies in face perception

    Perception 12:195–201.

    https://doi.org/10.1068/p120195

    • Google Scholar
32. 1. Fischl B
    2. Sereno MI
    3. Dale AM

    (1999a) Cortical surface-based analysis. II: Inflation, flattening, and a surface-based coordinate system

    Neuroimage 9:195–207.

    https://doi.org/10.1006/nimg.1998.0396

    • Google Scholar
33. 1. Fischl B
    2. Sereno MI
    3. Tootell RB
    4. Dale AM

    (1999b) High-resolution intersubject averaging and a coordinate system for the cortical surface

    Human Brain Mapping 8:272–284.

    https://doi.org/10.1002/(SICI)1097-0193(1999)8:43.0.CO;2-4

    • Google Scholar
34. 1. Fox MD
    2. Zhang D
    3. Snyder AZ
    4. Raichle ME

    (2009) The global signal and observed anticorrelated resting state brain networks

    Journal of Neurophysiology 101:3270–3283.

    https://doi.org/10.1152/jn.90777.2008

    • Google Scholar
35. 1. Gattass R
    2. Nascimento-Silva S
    3. Soares JG
    4. Lima B
    5. Jansen AK
    6. Diogo AC
    7. Farias MF
    8. Botelho MM
    9. Mariani OS
    10. Azzi J
    11. Fiorani M

    (2005) Cortical visual areas in monkeys: location, topography, connections, columns, plasticity and cortical dynamics

    Philosophical Transactions of the Royal Society of London Series B, Biological Sciences 360:709–731.

    https://doi.org/10.1098/rstb.2005.1629

    • Google Scholar
36. 1. Gattass R
    2. Sousa AP
    3. Mishkin M
    4. Ungerleider LG

    (1997) Cortical projections of area V2 in the macaque

    Cerebral Cortex 7:110–129.

    https://doi.org/10.1093/cercor/7.2.110

    • Google Scholar
37. 1. Ghuman AS
    2. van den Honert RN
    3. Martin A

    (2013) Interregional neural synchrony has similar dynamics during spontaneous and stimulus-driven states

    Scientific Reports 3:1481.

    https://doi.org/10.1038/srep01481

    • Google Scholar
38. 1. Goffaux V
    2. Hault B
    3. Michel C
    4. Vuong QC
    5. Rossion B

    (2005) The respective role of low and high spatial frequencies in supporting configural and featural processing of faces

    Perception 34:77–86.

    https://doi.org/10.1068/p5370

    • Google Scholar
39. 1. Goldberg JA
    2. Rokni U
    3. Sompolinsky H

    (2004) Patterns of ongoing activity and the functional architecture of the primary visual cortex

    Neuron 42:489–500.

    https://doi.org/10.1016/S0896-6273(04)00197-7

    • Google Scholar
40. 1. Gravel N
    2. Harvey B
    3. Nordhjem B
    4. Haak KV
    5. Dumoulin SO
    6. Renken R
    7. Curčić-Blake B
    8. Cornelissen FW

    (2014) Cortical connective field estimates from resting state fMRI activity

    Frontiers in Neuroscience 8:339.

    https://doi.org/10.3389/fnins.2014.00339

    • Google Scholar
41. 1. Haak KV
    2. Winawer J
    3. Harvey BM
    4. Renken R
    5. Dumoulin SO
    6. Wandell BA
    7. Cornelissen FW

    (2012) Connective field modeling

    Neuroimage 66:376–384.

    https://doi.org/10.1016/j.neuroimage.2012.10.037

    • Google Scholar
42. 1. Hansen KA
    2. Kay KN
    3. Gallant JL

    (2007) Topographic organization in and near human visual area V4

    The Journal of Neuroscience 27:11896–11911.

    https://doi.org/10.1523/JNEUROSCI.2991-07.2007

    • Google Scholar
43. 1. Harvey BM
    2. Dumoulin SO

    (2011) The relationship between cortical magnification factor and population receptive field size in human visual cortex: constancies in cortical architecture

    The Journal of Neuroscience 31:13604–13612.

    https://doi.org/10.1523/JNEUROSCI.2572-11.2011

    • Google Scholar
44. 1. Hasson U
    2. Levy I
    3. Behrmann M
    4. Hendler T
    5. Malach R

    (2002) Eccentricity bias as an organizing principle for human high-order object areas

    Neuron 34:479–490.

    https://doi.org/10.1016/S0896-6273(02)00662-1

    • Google Scholar
45. 1. Heinzle J
    2. Kahnt T
    3. Haynes JD

    (2011) Topographically specific functional connectivity between visual field maps in the human brain

    Neuroimage 56:1426–1436.

    https://doi.org/10.1016/j.neuroimage.2011.02.077

    • Google Scholar
46. 1. Henderson JM
    2. Hollingworth A

    (1999) High-level scene perception

    Annual Review of Psychology 50:243–271.

    https://doi.org/10.1146/annurev.psych.50.1.243

    • Google Scholar
47. 1. Honey CJ
    2. Sporns O
    3. Cammoun L
    4. Gigandet X
    5. Thiran JP
    6. Meuli R
    7. Hagmann P

    (2009) Predicting human resting-state functional connectivity from structural connectivity

    Proceedings of the National Academy of Sciences of USA 106:2035–2040.

    https://doi.org/10.1073/pnas.0811168106

    • Google Scholar
48. 1. Horton JC
    2. Hoyt WF

    (1991) The representation of the visual field in human striate cortex. A revision of the classic Holmes map

    Archives of Ophthalmology 109:816–824.

    https://doi.org/10.1001/archopht.1991.01080060080030

    • Google Scholar
49. 1. Hubel DH
    2. Wiesel TN

    (1967)

    Cortical and callosal connections concerned with the vertical meridian of visual fields in the cat

    Journal of Neurophysiology 30:1561–1573.

    • Google Scholar
50. 1. Jbabdi S
    2. Sotiropoulos SN
    3. Behrens TE

    (2013) The topographic connectome

    Current Opinion in Neurobiology 23:207–215.

    https://doi.org/10.1016/j.conb.2012.12.004

    • Google Scholar
51. 1. Jeffs J
    2. Ichida JM
    3. Federer F
    4. Angelucci A

    (2009) Anatomical evidence for classical and extra-classical receptive field completion across the discontinuous horizontal meridian representation of primate area V2

    Cerebral Cortex 19:963–981.

    https://doi.org/10.1093/cercor/bhn142

    • Google Scholar
52. 1. Kaas JH

    (1997) Topographic maps are fundamental to sensory processing

    Brain Research Bulletin 44:107–112.

    https://doi.org/10.1016/S0361-9230(97)00094-4

    • Google Scholar
53. 1. Keil MS

    (2008) Does face image statistics predict a preferred spatial frequency for human face processing?

    Proceedings Biological Sciences/ The Royal Society 275:2095–2100.

    https://doi.org/10.1098/rspb.2008.0486

    • Google Scholar
54. 1. Kennedy H
    2. Dehay C
    3. Bullier J

    (1986) Organization of the callosal connections of visual areas V1 and V2 in the macaque monkey

    The Journal of Comparative Neurology 247:398–415.

    https://doi.org/10.1002/cne.902470309

    • Google Scholar
55. 1. Kolster H
    2. Mandeville JB
    3. Arsenault JT
    4. Ekstrom LB
    5. Wald LL
    6. Vanduffel W

    (2009) Visual field map clusters in macaque extrastriate visual cortex

    The Journal of Neuroscience 29:7031–7039.

    https://doi.org/10.1523/JNEUROSCI.0518-09.2009

    • Google Scholar
56. 1. Leopold DA
    2. Murayama Y
    3. Logothetis NK

    (2003) Very slow activity fluctuations in monkey visual cortex: implications for functional brain imaging

    Cerebral Cortex 13:422–433.

    https://doi.org/10.1093/cercor/13.4.422

    • Google Scholar
57. 1. Levy I
    2. Hasson U
    3. Avidan G
    4. Hendler T
    5. Malach R

    (2001)

    Center-periphery organization of human object areas

    Nature Neuroscience 4:533–539.

    • Google Scholar
58. 1. Malach R
    2. Levy I
    3. Hasson U

    (2002) The topography of high-order human object areas

    Trends in Cognitive Sciences 6:176–184.

    https://doi.org/10.1016/S1364-6613(02)01870-3

    • Google Scholar
59. 1. Maunsell JH
    2. Van Essen DC

    (1983)

    The connections of the middle temporal visual area (MT) and their relationship to a cortical hierarchy in the macaque monkey

    The Journal of Neuroscience 3:2563–2586.

    • Google Scholar
60. 1. Murphy K
    2. Birn RM
    3. Handwerker DA
    4. Jones TB
    5. Bandettini PA

    (2009) The impact of global signal regression on resting state correlations: are anti-correlated networks introduced?

    Neuroimage 44:893–905.

    https://doi.org/10.1016/j.neuroimage.2008.09.036

    • Google Scholar
61. 1. Nakamura H
    2. Gattass R
    3. Desimone R
    4. Ungerleider LG

    (1993)

    The modular organization of projections from areas V1 and V2 to areas V4 and TEO in macaques

    The Journal of Neuroscience 13:3681–3691.

    • Google Scholar
62. 1. Nallasamy N
    2. Tsao DY

    (2011) Functional connectivity in the brain: effects of anesthesia

    The Neuroscientist 17:94–106.

    https://doi.org/10.1177/1073858410374126

    • Google Scholar
63. 1. Newsome WT
    2. Allman JM

    (1980) Interhemispheric connections of visual cortex in the owl monkey, Aotus trivirgatus, and the bushbaby, Galago senegalensis

    The Journal of Comparative Neurology 194:209–233.

    https://doi.org/10.1002/cne.901940111

    • Google Scholar
64. 1. Nir Y
    2. Hasson U
    3. Levy I
    4. Yeshurun Y
    5. Malach R

    (2006) Widespread functional connectivity and fMRI fluctuations in human visual cortex in the absence of visual stimulation

    Neuroimage 30:1313–1324.

    https://doi.org/10.1016/j.neuroimage.2005.11.018

    • Google Scholar
65. 1. Nir Y
    2. Mukamel R
    3. Dinstein I
    4. Privman E
    5. Harel M
    6. Fisch L
    7. Gelbard-Sagiv H
    8. Kipervasser S
    9. Andelman F
    10. Neufeld MY
    11. Kramer U
    12. Arieli A
    13. Fried I
    14. Malach R

    (2008) Interhemispheric correlations of slow spontaneous neuronal fluctuations revealed in human sensory cortex

    Nature Neuroscience 11:1100–1108.

    https://doi.org/10.1038/nn.2177

    • Google Scholar
66. 1. Oliva A
    2. Schyns PG

    (1997) Coarse blobs or fine edges? Evidence that information diagnosticity changes the perception of complex visual stimuli

    Cognitive Psychology 34:72–107.

    https://doi.org/10.1006/cogp.1997.0667

    • Google Scholar
67. 1. Oliva A
    2. Torralba A

    (2006) Building the gist of a scene: the role of global image features in recognition

    Progress in Brain Research 155:23–36.

    https://doi.org/10.1016/S0079-6123(06)55002-2

    • Google Scholar
68. 1. Parkes LM
    2. Schwarzbach JV
    3. Bouts AA
    4. Deckers RH
    5. Pullens P
    6. Kerskens CM
    7. Norris DG

    (2005) Quantifying the spatial resolution of the gradient echo and spin echo BOLD response at 3 Tesla

    Magnetic Resonance in Medicine 54:1465–1472.

    https://doi.org/10.1002/mrm.20712

    • Google Scholar
69. 1. Raemaekers M
    2. Schellekens W
    3. Van Wezel RJ
    4. Petridou N
    5. Kristo G
    6. Ramsey NF

    (2014) Patterns of resting state connectivity in human primary visual cortical areas: a 7T fMRI study

    Neuroimage 84:911–921.

    https://doi.org/10.1016/j.neuroimage.2013.09.060

    • Google Scholar
70. 1. Rosa MG

    (2002) Visual maps in the adult primate cerebral cortex: some implications for brain development and evolution

    Brazilian Journal of Medical and Biological Research 35:1485–1498.

    https://doi.org/10.1590/S0100-879X2002001200008

    • Google Scholar
71. 1. Rosa MG
    2. Palmer SM
    3. Gamberini M
    4. Burman KJ
    5. Yu HH
    6. Reser DH
    7. Bourne JA
    8. Tweedale R
    9. Galletti C

    (2009) Connections of the dorsomedial visual area: pathways for early integration of dorsal and ventral streams in extrastriate cortex

    The Journal of Neuroscience 29:4548–4563.

    https://doi.org/10.1523/JNEUROSCI.0529-09.2009

    • Google Scholar
72. 1. Rosa MG
    2. Tweedale R

    (2005) Brain maps, great and small: lessons from comparative studies of primate visual cortical organization

    Philosophical Transactions of the Royal Society of London Series B, Biological Sciences 360:665–691.

    https://doi.org/10.1098/rstb.2005.1626

    • Google Scholar
73. 1. Saad ZS
    2. Gotts SJ
    3. Murphy K
    4. Chen G
    5. Jo HJ
    6. Martin A
    7. Cox RW

    (2012) Trouble at rest: how correlation patterns and group differences become distorted after global signal regression

    Brain Connectivity 2:25–32.

    https://doi.org/10.1089/brain.2012.0080

    • Google Scholar
74. 1. Schira MM
    2. Tyler CW
    3. Breakspear M
    4. Spehar B

    (2009) The foveal confluence in human visual cortex

    The Journal of Neuroscience 29:9050–9058.

    https://doi.org/10.1523/JNEUROSCI.1760-09.2009

    • Google Scholar
75. 1. Sereno MI
    2. Dale AM
    3. Reppas JB
    4. Kwong KK
    5. Belliveau JW
    6. Brady TJ
    7. Rosen BR
    8. Tootell RB

    (1995) Borders of multiple visual areas in humans revealed by functional magnetic resonance imaging

    Science 268:889–893.

    https://doi.org/10.1126/science.7754376

    • Google Scholar
76. 1. Shmuel A
    2. Yacoub E
    3. Chaimow D
    4. Logothetis NK
    5. Ugurbil K

    (2007) Spatio-temporal point-spread function of fMRI signal in human gray matter at 7 Tesla

    Neuroimage 35:539–552.

    https://doi.org/10.1016/j.neuroimage.2006.12.030

    • Google Scholar
77. 1. Silver MA
    2. Kastner S

    (2009) Topographic maps in human frontal and parietal cortex

    Trends in Cognitive Sciences 13:488–495.

    https://doi.org/10.1016/j.tics.2009.08.005

    • Google Scholar
78. 1. Smith SM
    2. Miller KL
    3. Moeller S
    4. XU J
    5. Auerbach EJ
    6. Woolrich MW
    7. Beckmann CF
    8. Jenkinson M
    9. Andersson J
    10. Glasser MF
    11. Van Essen DC
    12. Feinberg DA
    13. Yacoub ES
    14. Ugurbil K

    (2012) Temporally-independent functional modes of spontaneous brain activity

    Proceedings of the National Academy of Sciences of USA 109:3131–3136.

    https://doi.org/10.1073/pnas.1121329109

    • Google Scholar
79. 1. Sporns O
    2. Honey CJ

    (2013) Topographic dynamics in the resting brain

    Neuron 78:955–956.

    https://doi.org/10.1016/j.neuron.2013.05.037

    • Google Scholar
80. 1. Stepniewska I
    2. Kaas JH

    (1996) Topographic patterns of V2 cortical connections in macaque monkeys

    The Journal of Comparative Neurology 371:129–152.

    https://doi.org/10.1002/(SICI)1096-9861(19960715)371:13.0.CO;2-5

    • Google Scholar
81. 1. Swisher JD
    2. Halko MA
    3. Merabet LB
    4. McMains SA
    5. Somers DC

    (2007) Visual topography of human intraparietal sulcus

    The Journal of Neuroscience 27:5326–5337.

    https://doi.org/10.1523/JNEUROSCI.0991-07.2007

    • Google Scholar
82. 1. Tsodyks M
    2. Kenet T
    3. Grinvald A
    4. Arieli A

    (1999) Linking spontaneous activity of single cortical neurons and the underlying functional architecture

    Science 286:1943–1946.

    https://doi.org/10.1126/science.286.5446.1943

    • Google Scholar
83. 1. Ungerleider LG
    2. Galkin TW
    3. Desimone R
    4. Gattass R

    (2008) Cortical connections of area V4 in the macaque

    Cerebral Cortex 18:477–499.

    https://doi.org/10.1093/cercor/bhm061

    • Google Scholar
84. 1. Ungerleider LG
    2. Galkin TW
    3. Desimone R
    4. Gattass R

    (2014) Subcortical projections of area v2 in the macaque

    Journal of Cognitive Neuroscience 26:1220–1233.

    https://doi.org/10.1162/jocn_a_00571

    • Google Scholar
85. 1. Van Essen DC
    2. Zeki SM

    (1978) The topographic organization of rhesus monkey prestriate cortex

    The Journal of Physiology 277:193–226.

    https://doi.org/10.1113/jphysiol.1978.sp012269

    • Google Scholar
86. 1. Van Essen DC
    2. Newsome WT
    3. Maunsell JH

    (1984) The visual field representation in striate cortex of the macaque monkey: asymmetries, anisotropies, and individual variability

    Vision Research 24:429–448.

    https://doi.org/10.1016/0042-6989(84)90041-5

    • Google Scholar
87. 1. Vincent JL
    2. Patel GH
    3. Fox MD
    4. Snyder AZ
    5. Baker JT
    6. Van Essen DC
    7. Zempel JM
    8. Snyder LH
    9. Corbetta M
    10. Raichle ME

    (2007) Intrinsic functional architecture in the anaesthetized monkey brain

    Nature 447:83–86.

    https://doi.org/10.1038/nature05758

    • Google Scholar
88. 1. Wandell BA
    2. Dumoulin SO
    3. Brewer AA

    (2007) Visual field maps in human cortex

    Neuron 56:366–383.

    https://doi.org/10.1016/j.neuron.2007.10.012

    • Google Scholar
89. 1. Wandell BA
    2. Winawer J

    (2011) Imaging retinotopic maps in the human brain

    Vision Research 51:718–737.

    https://doi.org/10.1016/j.visres.2010.08.004

    • Google Scholar
90. 1. Wang L
    2. Mruczek RE
    3. Arcaro MJ
    4. Kastner S

    (2014)

    Probabilistic maps of visual topography in human cortex

    Cerebral Cortex.

    • Google Scholar
91. 1. Wang Z
    2. Chen LM
    3. Negyessy L
    4. Friedman RM
    5. Mishra A
    6. Gore JC
    7. Roe AW

    (2013) The relationship of anatomical and functional connectivity to resting-state connectivity in primate somatosensory cortex

    Neuron 78:1116–1126.

    https://doi.org/10.1016/j.neuron.2013.04.023

    • Google Scholar
92. 1. Wig GS
    2. Laumann TO
    3. Petersen SE

    (2014) An approach for parcellating human cortical areas using resting-state correlations

    Neuroimage 93:276–291.

    https://doi.org/10.1016/j.neuroimage.2013.07.035

    • Google Scholar
93. Book

    1. Yarbus AL

    (1967)

    Eye movements and vision

    New York: Plenum.

    • Google Scholar
94. 1. Yeo BT
    2. Krienen FM
    3. Sepulcre J
    4. Sabuncu MR
    5. Lashkari D
    6. Hollinshead M
    7. Roffman JL
    8. Smoller JW
    9. Zollei L
    10. Polimeni JR
    11. Fischl B
    12. Liu H
    13. Buckner RL

    (2011) The organization of the human cerebral cortex estimated by intrinsic functional connectivity

    Journal of Neurophysiology 106:1125–1165.

    https://doi.org/10.1152/jn.00338.2011

    • Google Scholar
95. 1. Zeki SM

    (1969) The secondary visual areas of the monkey

    Brain Research 13:197–226.

    https://doi.org/10.1016/0006-8993(69)90282-0

    • Google Scholar
96. 1. Zeki SM

    (1971) Cortical projections from two prestriate areas in the monkey

    Brain Research 34:19–35.

    https://doi.org/10.1016/0006-8993(71)90348-9

    • Google Scholar

Author details

1. Michael J Arcaro

   1. Princeton Neuroscience Institute, Princeton University, Princeton, United States
   2. Department of Psychology, Princeton University, Princeton, United States

   Contribution

   MJA, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   marcaro@princeton.edu

   Competing interests

   The authors declare that no competing interests exist.

2. Christopher J Honey

   Department of Psychology, University of Toronto, Toronto, Canada

   Contribution

   CJH, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

3. Ryan EB Mruczek

   Department of Psychology, Worcester State University, Worcester, United States

   Contribution

   REBM, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

4. Sabine Kastner

   1. Princeton Neuroscience Institute, Princeton University, Princeton, United States
   2. Department of Psychology, Princeton University, Princeton, United States

   Contribution

   SK, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

5. Uri Hasson

   1. Princeton Neuroscience Institute, Princeton University, Princeton, United States
   2. Department of Psychology, Princeton University, Princeton, United States

   Contribution

   UH, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

• 2,482

  views

• 395

  downloads

• 43

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.