We rely on our sense of vision to perceive the world around us and the objects within it. We also use vision to guide our interactions with objects. One of the most influential theories in cognitive neuroscience is the idea that separate pathways within the brain support these two processes. The ventral pathway is in charge of vision-for-perception. It analyses the features that help us recognize objects, such as their color, size or shape, enabling us to identify the hammer in a toolbox, for example. The dorsal pathway is responsible for vision-for-action. It processes features that help us interact with objects, such as their movement and location, enabling us to use the hammer to strike a nail.

However, recent studies have suggested that the ventral and dorsal pathways may not be as independent as originally thought. Freud et al. now test this idea by examining if the dorsal vision-for-action pathway can also perceive and process objects.

Healthy volunteers viewed pictures of objects while lying inside a brain scanner. Some of the objects in the pictures were intact, whereas others had been distorted. If a brain region shows greater activation when viewing intact objects than distorted ones, it implies that that region is sensitive to the normal shapes of objects. Freud et al. found that both the ventral and dorsal pathways were sensitive to shape, with some areas in the two pathways showing highly similar responses. Furthermore, the shape sensitivity of certain regions within the dorsal pathway correlated with the volunteers’ ability to recognize the objects. This suggests that regions distributed across both pathways – and not just the ventral one – may contribute to object recognition.

The two-pathways hypothesis has governed our understanding of vision and of other sensory systems including hearing for several decades. By challenging the binary distinction between the two pathways, the results of Freud et al. suggest that models of sensory processing may require updating. This improved understanding may ultimately improve diagnosis and treatment of perceptual disorders such as agnosia, in which patients struggle to recognize objects.

Shape is the most fundamental perceived property of objects, and, accordingly, shape processing is crucial for successful visual object recognition (Palmer, 1999). Deriving information about the shapes of objects in the input has long been considered to be under the purview of the ventral visual pathway (Goodale and Milner, 1992; Ungerleider and Mishkin, 1982) and many functional imaging studies have provided evidence to support this claim (e.g., Freud et al., 2013; Grill-Spector et al., 1998; Kourtzi and Kanwisher, 2000; Lerner et al., 2001; Malach et al., 1995). Specifically, these investigations have uncovered a gradient of shape sensitivity in the ventral pathway, with posterior regions in the early to mid-visual cortex (i.e., V1-hV4) being less responsive to the shape of the object than more anterior regions, such as the Lateral Occipital Cortex (LOC) and the Fusiform Gyrus (FG). Additionally, shape sensitivity in these latter, anterior regions is correlated with perceptual abilities, and a lesion to these areas results in an impairment in object perception (e.g., Freud et al., 2017a; Goodale et al., 1991; Konen et al., 2011).

It appears, however, that shape perception is not solely a product of the computations mediated by the ventral pathway. Human neuroimaging studies (Freud et al., 2015; Jeong and Xu, 2016; Konen and Kastner, 2008; Zachariou et al., 2014; Zachariou et al., 2017; Bracci and Op de Beeck, 2016; Bracci et al., 2017), primate electrophysiological studies (Durand et al., 2007; Janssen et al., 2008; Janssen et al., 2000; Van Dromme et al., 2016); for review see Theys et al., 2015), comparative studies between human and primates (Denys et al., 2004; Sawamura et al., 2005) and human neuropsychology (Freud et al., 2017a) have all uncovered object representations mediated by the dorsal pathway, even under conditions in which no visuomotor response is required. However, many questions remain concerning the topographical organization of the dorsal object representations, the nature of the visual properties encoded in these representations and their contribution to perceptual behavior.

In the current study, we addressed these questions in two studies, each combining functional magnetic resonance imaging (fMRI) and psychophysical measures. We hypothesized that the large-scale topographical organization of dorsal object representations obeys a spatial gradient in which posterior regions of the pathway derive shape representations and contribute to visual perception, whereas more anterior regions derive representations that are better tuned to subserve visuomotor behaviors (Freud et al., 2016). Our motivation stems from two key issues concerning dorsal pathway function: the first is the need to reconcile recent findings of dorsal pathway activation under conditions of perception (as noted above) with the well-established findings that the dorsal pathway is engaged in object-directed actions such as grasping and manipulation (e.g., Culham et al., 2003; Fabbri et al., 2016; for a review, see Gallivan and Culham, 2015). The second concerns the connectivity that anchors the parietal representational continuum at one end with more caudal regions connected to visual cortex (Greenberg et al., 2012) and, at the other end, with more rostral regions anatomically and functionally connected to motor cortex (Davare et al., 2010).

We adopt an approach that has been used previously to characterize the neural basis of shape processing in ventral cortex in response to stimuli in which shape information is parametrically eliminated by increasingly distorting images of objects (Lerner et al., 2001) (see Figure 1A for example). The decrease in BOLD activation with increased scrambling serves as an index of shape sensitivity (Denys et al., 2004; Grill-Spector et al., 1998; Lerner et al., 2001; Malach et al., 1995; Murray et al., 2002).

Figure 1

Download asset Open asset

Following these previous studies, in the first experiment, we adopted a parametric box scrambling procedure to map the large-scale organization of shape processing along the ventral and dorsal visual pathways and the contribution of these representations to shape perception. In addition, we employed more advanced multivariate analyses to further elucidate the nature of representations and the similarities or differences in these representations in the two pathways. However, the box-scrambling method has some inherent limitations. First, it increases the number of edges, and early visual cortex is especially sensitive to this type of information (Grill-Spector et al., 1998; Lerner et al., 2001). Second, box scrambling disrupts crucial shape attributes such as good continuation and figure-ground segregation (Koffka, 1935; Qiu and von der Heydt, 2005; Read et al., 1997), and, unsurprisingly, eliminates the identity information of the stimulus. Furthermore, in the box scrambling experiment, participants were shown each object at every level of scrambling and so there may have been priming or adaptation for the same object across levels of scrambling.

To circumvent these limitations and verify the results, in the second experiment we used a diffeomorphic scrambling manipulation that increasingly precludes the identification of the object but preserves some fundamental shape properties by repeatedly applying a flow field generated from a set of two-dimensional cosine components with random phase and amplitude (Stojanoski and Cusack, 2014). Consequently, some shape information (i.e., the presence of an object or figure that is clearly differentiated from the background and has well-defined boundaries, relatively uniform texture and center of mass; see methods and Figure 1B for quantitative analyses of object distortion) is still largely available even in the most distorted version, but the object itself is parametrically distorted and unrecognizable. This manipulation enabled us to re-examine the large-scale organization of the two pathways, and to disentangle sensitivity to shape versus identity information. Last, each object was presented at only one level of distortion (counterbalanced across participants) to remove possible priming effects.

The large-scale organization of shape processing along the dorsal and ventral visual pathways was assessed in two imaging experiments, by computing the level of the BOLD signal as a function of the decrease in shape information across five levels. Three analytical approaches were employed: a novel voxel-wise approach that provides a continuous voxel-wise mapping of shape processing along the two pathways, a more traditional ROI analysis, and a multivariate representational similarity analysis (RSA) to uncover additional information about the similarity of spatial patterns of activation as a function of scrambling.

Univariate analysis

Box scrambling experiment

To map the topographical organization of shape processing along the two pathways in a continuous fashion, we generated a voxel-wise map of shape sensitivity for all visually selective voxels. In each voxel, beta weights were extracted for each of the five stimulus conditions from most scrambled to intact (S4, S3, S2, S1, Full), and the linear slope between these conditions was used as an index of shape sensitivity. Voxels that are shape sensitive were those in which activation, as a function of object coherence, had a positive slope. Conversely, voxels that showed greater sensitivity for scrambled than intact images had a negative slope. To map the profile of shape sensitivity across all voxels, we then applied a piecewise linear regression across all the voxels using two (rather than one) linear components (see Methods for details). In each pathway, the piecewise linear regression increased the R² when compared with a simple linear regression (ps <0.05), thereby providing a better fit to the data than a one-component linear model (see Materials and methods).

In the ventral pathway (Figure 2A), a negative slope was found in the vicinity of the calcarine sulcus and in the posterior occipital lobe more generally, and likely reflects the sensitivity of early visual cortex to edges and high spatial frequency information in the input (Lerner et al., 2001); for an additional interpretation, see Murray et al., 2002). In more anterior regions of the ventral pathway, a positive slope was evident both on the lateral and the inferior surface of the occipitotemporal cortex (i.e., LO, Parahippocampal gyrus and Fusiform gyrus). However, in even more anterior regions (i.e., anterior and medial temporal cortex), a decrease in shape sensitivity was detected. The piecewise regression revealed that the data were well described by two linear components. The first component reflected the increase in shape sensitivity in the transition between early visual cortex and object-selective cortex and was characterized by a positive correlation between slope and location on the posterior-anterior axis [LH: t₍₁₀₎=8.21, q < 0.00001 CI {0.39, 0.69}; RH: t₍₁₀₎=12.1 q < 0.00001, CI {0.43, 0.62}]. The second component was characterized by a decrease in shape sensitivity, as evident from the negative correlation between location on the posterior-anterior axis and slope [LH: t₍₁₀₎=9.5, q = 0.00001, CI{−0.32, −0.51}; RH: t₍₁₀₎=6.94, q = 0.00001, CI {−0.20 −0.39}; Figure 2B]. The robust positive first component is consistent with previous investigations of the ventral pathway (Lerner et al., 2001; Murray et al., 2002), and replicates the hierarchical nature of shape processing in which more elaborate and complex representations are derived as one moves rostrally (Grill-Spector and Weiner, 2014). The reduction in shape sensitivity in the anterior parts of the temporal lobe might reflect a transition to more semantic and memory-based representation (Kravitz et al., 2013; Visser et al., 2010), rather than shape sensitivity per se, but might also be a consequence of MRI distortion effects which are more common in these regions (Olman et al., 2009).

Figure 2 with 2 supplements see all

Download asset Open asset

Figure 2—source data 1

Individual data points for Figure 2C.

https://doi.org/10.7554/eLife.27576.007

Download elife-27576-fig2-data1-v2.xlsx

Figure 2—source data 2

Individual data points for Figure 2D.

https://doi.org/10.7554/eLife.27576.008

Download elife-27576-fig2-data2-v2.xlsx

Figure 2—source data 3

Individual data points for Figure 2E.

https://doi.org/10.7554/eLife.27576.009

Download elife-27576-fig2-data3-v2.xlsx

Figure 2—source data 4

Individual data points for Figure 2—figure supplement 1 (upper panel).

https://doi.org/10.7554/eLife.27576.010

Download elife-27576-fig2-data4-v2.xlsx

Figure 2—source data 5

Individual data points for Figure 2—figure supplement 1 (lower panel).

https://doi.org/10.7554/eLife.27576.011

Download elife-27576-fig2-data5-v2.xlsx

Figure 2—source data 6

Individual data points for Figure 2—figure supplement 2.

https://doi.org/10.7554/eLife.27576.012

Download elife-27576-fig2-data6-v2.xlsx

As in the ventral pathway, shape sensitivity in the dorsal pathway was not evident in early visual areas, but emerged in extrastriate occipital cortex and reached a peak in the posterior intraparietal sulcus (IPS). Consistent with the findings of the ventral pathway, shape selectivity gradually decreased from the posterior parietal lobe to more anterior regions (Figure 2A). Accordingly, the piecewise regression revealed a positive correlation for the first component, which reflects the increase in shape sensitivity [LH: t₍₁₀₎=5.15, q = 0.0004 CI {0.32, 0.82}; RH: t₍₁₀₎=10, q = 0.00001, CI {0.62, 0.97}], and a robust negative correlation for the second component, which reflects the decrease in shape sensitivity in the anterior parts of the dorsal pathway [LH: t₍₁₀₎=8.54, q = 0.00001 CI {−0.32, −0.55}; RH: t₍₁₀₎=8.94, q = 0.000001, CI {−0.34, −0.57}; Figure 2C].

The key finding here is that the two pathways have similar topographical organization. In addition, these findings provide novel evidence for the nature of shape processing in the dorsal pathway and indicate that object representations in this pathway are not monolithic but, rather, differ qualitatively along the posterior-anterior axis (Freud et al., 2016).

The mapping of topographical organization of shape sensitivity was done along the posterior-anterior axis. However, given the curvature of the two pathways in the brain, it might be the case that shape information is also modulated by the inferior-superior axis. To examine this possibility, we calculated the distance between each voxel to the most posterior voxel using both Y and Z coordinates and then evaluated the shape-sensitivity gradient. This analysis replicates the organization revealed using only the Y-axis and indicates the Y-axis serves as the key dimension (see Figure 2—figure supplement 1, lower panel).

Despite the apparent qualitative similarity in the large-scale topographical organization of shape processing along the two pathways, further analyses also revealed some differences between the ventral and dorsal pathways. In particular, the maximal shape sensitivity (the apex or inflection point of the regression line) was more posterior in the dorsal relative to the ventral pathway [F_(1,10) = 22.11, p=0.0008, η_p² = 0.68, CI {12 21}; Figure 2D]. This effect was reproduced when the main dependent variable was the distance from the most posterior voxel based on the Y and Z coordinates [F_(1,10) = 91, p=0.000002, η_p² = 0.90, CI {16 27}].

Additionally, the average sensitivity of all shape-selective voxels (slope > 0) was significantly greater in the ventral than dorsal pathway [F_(1,10) = 45, p=0.00005, η_p² = 0.81, CI {0.01 0.02}; Figure 2E]. Interestingly, this main effect was modulated by an interaction with hemisphere [F_(1,10) = 11.37, p=0.007, η_p² = 0.53, CI {0.01 0.02}] with greater difference between the pathways in the left than the right hemisphere. The greater difference between the pathways in shape sensitivity in the left hemisphere might reflect hemispheric specialization, such that the left dorsal pathway is more tuned to visuomotor aspects than the right dorsal pathway.

Finally, in addition to the continuous analysis described above, the data were analyzed using an ROI approach. ROIs were defined based on a probabilistic atlas (Wang et al., 2015) and the aIPS ROI was defined based on a meta-analysis (Neurosynth.org). This analysis reproduced the results reported above and demonstrated the two components organization of the two pathways (Figure 2—figure supplement 2).

Diffeomorphic scrambling experiment

The second experiment utilized the diffeomorphic stimuli to replicate the large-scale organization observed for the box scrambling method as well as to provide further insight into the nature of representations derived by the different regions.

In the ventral pathway, the piecewise regression revealed a positive correlation between slope and location on the posterior-anterior axis (first component), which reflects the emergence of shape sensitivity in the transition from early visual cortex to object selective cortex [LH: t₍₁₀₎=25.2 q<0.0000001, CI [0.59, 0.70]; RH: t₍₁₀₎=13.2, q = 0.000004, CI {44, 0.63}], and a negative correlation for the second component, which reflects the decrease in shape sensitivity in the most anterior parts of the temporal lobe [LH: t₍₁₀₎= 9.77, q = 0.0000005, CI{−0.25, −0.40}; RH: t₍₁₀₎=4.57, q = 0.001, CI{−0.09, −0.27}].

These findings replicate those of the box scrambling experiment, in which sensitivity to shape information was not present in posterior regions of the ventral pathway, increased in more rostral parts of the pathway and then decreased in the more anterior temporal regions. Interestingly, however, on the lateral surface of the ventral pathway (corresponding to LOC, see also ROI analysis, Figure 3—figure supplement 2), a flatter slope was found (i.e., similar activation to the different levels of image distortion), suggesting that this region was less sensitive to the diffeomorphic manipulation than to the box scrambling manipulation. Consistent with previous reports (Malach et al., 1995; Margalit et al., 2016), LOC appears to represent the presence of a well-defined shape, with clear contours, rather than a representation of the object’s identity (Figure 3A).

Figure 3 with 2 supplements see all

Download asset Open asset

Figure 3—source data 1

Individual data points for Figure 3C.

https://doi.org/10.7554/eLife.27576.016

Download elife-27576-fig3-data1-v2.xlsx

Figure 3—source data 2

Individual data points for Figure 3D.

https://doi.org/10.7554/eLife.27576.017

Download elife-27576-fig3-data2-v2.xlsx

Figure 3—source data 3

Individual data points for Figure 3E.

https://doi.org/10.7554/eLife.27576.018

Download elife-27576-fig3-data3-v2.xlsx

Figure 3—source data 4

Individual data points for Figure 3—figure supplement 1 (upper panel).

https://doi.org/10.7554/eLife.27576.019

Download elife-27576-fig3-data4-v2.xlsx

Figure 3—source data 5

Individual data points for Figure 3—figure supplement 1 (lower panel).

https://doi.org/10.7554/eLife.27576.020

Download elife-27576-fig3-data5-v2.xlsx

Figure 3—source data 6

Individual data points Figure 3—figure supplement 2.

https://doi.org/10.7554/eLife.27576.021

Download elife-27576-fig3-data6-v2.xlsx

The piecewise regression of the dorsal pathway revealed two opposite linear components. The first component was characterized by a positive correlation between slope and location on the y-axis, which reflects the transition from early visual cortex to object-selective cortex [LH: t ₍₁₀₎=3.98, q = 0.002, CI {0.22, 0.80}; RH: t ₍₁₀₎=7.31, q = 0.00005, CI{43, 0.81}]. In contrast, the second component, which reflects the representational gradient along the IPS, was characterized by a negative correlation between slope and location of the voxels on the posterior-anterior axis [LH: t ₍₁₀₎=5.96, q = 0.0001, CI {−0.16, −0.35}; RH: t ₍₁₀₎=6.39, q = 0.0001, CI{−0.19, −0.39}].

A repeated measures ANOVA on the correlation values (between slope and location of voxels on the y-axis), with Experiment as a between-subjects variable, and pathway, component and hemisphere as within-subject variables, revealed no interactions between Experiment and the other variables [all ps > 0.05]. This result indicates that, despite the difference in the overall shape sensitivity between the two experiments, they provided consistent information on the overall large-scale organization of shape sensitivity.

As in box scrambling experiment, the maximal shape sensitivity emerged in more posterior regions of the dorsal pathway compared with the ventral pathway [F _(1,10)=18.78, p=0.001, η_p² = 0.65], and the effect held when distance based on the Y and Z axes served as the dependent variable [F _(1,10)=45, p=0.00005, η_p² = 0.81]. The similarity between the two experiments was confirmed by an ANOVA that included Experiment as a between-subjects variable. In particular, a main effect of pathway was found [F _(1,20)=56, p=0.000001, η_p² = 0.73] and was not modulated by experiment [F _(1,20) < 1]. Interestingly, the inflection point of the regression was more anterior in the diffeomorphic scrambling experiment compared with box scrambling experiment [F _(1,20)=8.45, p=0.008, η_p² = 0.29], and this might reflect the increased high-level shape or identity processing associated with the diffeomorphic scrambling stimuli.

In addition, we compared the overall neural sensitivity to shape information in all shape-selective voxels (slope > 0). The average shape sensitivity was greater in the ventral than the dorsal pathway [F _(1,10)=37, p=0.0001, η_p² = 0.79]. The comparison between the two experiments revealed two main effects, with no interaction. As expected, there was a main effect of pathway: shape sensitivity was greater overall in the ventral than dorsal pathway across the two experiments [F _(1,20)=83, p=0.000001, η_p² = 0.8]. There was also an additional main effect of Experiment, with greater average positive slope in the box scrambling experiment than the diffeomorphic scrambling experiment, reflecting the different nature of the two manipulations [F _(1,20)=32, p=0.00001, η_p² = 0.62; see also Figure 1]. Notably, an interaction between pathway and hemisphere [F _(1,20)=10.85, p=0.003, η_p² = 0.35] was found and was not modulated by experiment [F _(1,20)<1]. This interaction was the result of a greater difference between the two pathways in the left than right hemisphere, and therefore further implicates hemispheric differences in shape processing.

Finally, the ROI analysis (Figure 3—figure supplement 2) confirms that the large-scale organization of object processing was largely replicated in the diffeomorphic scrambling experiment. In particular, in the dorsal pathway, shape sensitivity was first evident in V3b, reached a peak at IPS0/V3b and then decreased, and was not evident in the more anterior IPS4 and aIPS, confirming the posterior-anterior representational gradient. Note that the ROIs-based analysis did not permit an exploration of the ventral pathway second component, since this component was found to be in more anterior parts of the ventral pathway in the diffeomorphic scrambling experiment and the probabilistic atlas does not include these more anterior regions.

Representational similarity analysis

Box scrambling experiment

So far, we have presented results based on a univariate approach. Despite the robustness of this method and the convergence of the voxel-wise and ROI approaches, these analyses only consider the magnitude but not the pattern of the fMRI activation. Recent investigations employing multivariate approaches have successfully uncovered important features of the neural representations of objects in both the dorsal and the ventral pathways (e.g., Bracci et al., 2017; Bracci and Op de Beeck, 2016; Fabbri et al., 2016).

To extend our results, we utilized a multivariate approach of representational similarity analysis (Kriegeskorte et al., 2008; RSA). Notably, most previous investigations utilize the RSA approach to uncover the representational space of individual exemplars. The RSA employed here, however, is unlike the standard RSA procedure. Given that the present study used a block design to increase the statistical power in the service of the voxel-by-voxel univariate mapping described above, we could not evaluate the representation of individual exemplars. Instead, the RSA was applied at the level of a block (where a single block contains intact or scrambled objects to differing degrees). This procedure still allows us to explore how the availability of shape information modulated the representational content of different ROIs.

The RSA computed the similarity of spatial activation patterns across conditions for different ROIs and then tested whether these data could be accounted for by a model in which representations of images will be more similar to each other if they possess similar levels of shape information. For example, in shape selective regions, the response pattern to intact images is predicted to be more correlated with the response pattern to images scrambled to four pieces compared with the response pattern to images scrambled to 64 pieces. Correspondingly, response patterns for the 256 scrambled inputs should be more correlated with voxels that respond to the 64 scrambled inputs versus intact objects (Figure 4A).

Figure 4

Download asset Open asset

Figure 4—source data 1

Individual data points for Figure 4B.

https://doi.org/10.7554/eLife.27576.023

Download elife-27576-fig4-data1-v2.xlsx

Figure 4—source data 2

Individual data points for Figure 4D.

https://doi.org/10.7554/eLife.27576.024

Download elife-27576-fig4-data2-v2.xlsx

The results of the RSA are depicted in Figure 4B. Both early visual cortices and object-selective cortices were correlated with the shape model. The correlation of early visual cortices to the shape model can be explained by the negative slope observed in those regions that carry information about the presence/absence of a shape and about low-level similarities (Bracci and Op de Beeck, 2016). More importantly, and compatible with the univariate analysis, mid-posterior parts of the dorsal pathway and ventral pathways were highly correlated with the shape sensitivity model, and this sensitivity decreased in more anterior parts of the parietal cortex and temporal lobe.

Finally, RSA also permits a second-level analysis in which multidimensional scaling is used to place ROIs into a representational similarity space (here two-dimensional) such that those that are similar in representational structure are in closer proximity and those in which representational structure is dissimilar are more distant (Figure 4B) (see Materials and methods). This visualization further demonstrates the separation between early visual cortices and object-selective cortices. More importantly, this analysis also reveals the association in the representational structures of the two pathways. In particular, the patterns of activation in the posterior parts of the dorsal pathway (i.e. V3a-IPS1) were highly correlated with regions of the lateral ventral pathway (i.e., LO1, LO2, and TO1).

To quantify the representational similarity within and between pathways, we compared the correlations of the object-selective ROIs within each pathway (i.e. lateral-ventral ROIs to inferior-ventral ROIs and posterior-dorsal ROIs to anterior-dorsal ROIs) to the correlations of the object-selective ROIs between the two pathways (i.e., posterior-dorsal ROIs with lateral-ventral ROIs). Interestingly, this analysis revealed that the between-pathways correlations were higher than the within-pathway correlations both in the ventral [F _(1,10)=26.1, p=0.0004, η_p² = 0.72,, CI{0.05, 0.13}] and dorsal [F _(1,10)=19.3, p=0.001, η_p² = 0.65, CI{0.05, 0.16}] pathways, further challenging a binary distinction between the two pathways. Notably, this effect is not the result of the assignment of a particular ROI to the posterior/anterior dorsal pathway, and was replicated even when aIPS was excluded from the analysis, or when IPS2 was included in the posterior (rather than anterior) dorsal pathway.

The RSA results converge with the univariate results and demonstrate that regions in the dorsal and ventral pathway share highly similar representational structures based on the availability of shape information, thereby challenging the binary distinction between the two pathways.

Diffeomorphic scrambling experiment

The results of the RSA are depicted in Figure 5B. Both early visual cortices and object-selective cortices were correlated with the shape model, however, these correlations were significantly reduced in comparison to the box scrambling experiment (for a direct comparison see Figure 5—figure supplement 1). Importantly, this reduction is not an artifact of noise in the data as evident from the noise ceiling calculation (gray bars, Figure 5B). The reduction in the correlation to the shape model is therefore more likely related to the fact that the impact of shape distortion was reduced for the diffeomorphic stimuli. Importantly however, the large-scale organization of the dorsal pathway (particularly in the right hemisphere) was replicated in the diffeomorphic scrambling experiment, as the highest correlations with the shape model were observed in V3a-b and IPS0.

Figure 5 with 1 supplement see all

Download asset Open asset

Figure 5—source data 1

Individual data points for Figure 5B.

https://doi.org/10.7554/eLife.27576.027

Download elife-27576-fig5-data1-v2.xlsx

Figure 5—source data 2

Individual data points for Figure 5D.

https://doi.org/10.7554/eLife.27576.028

Download elife-27576-fig5-data2-v2.xlsx

A robust reduction in the correlation to the shape model was observed along the ventral pathway, particularly along its lateral surface. This was confirmed by an ANOVA that revealed an interaction between experiment and ROI [F_(10,200) = 14.48, p=0.0000001, η_p² =.42]. The robust decrease of correlation to the shape model, further suggests that the LOC codes shape information rather than a high-level abstraction of the object’s identity.

The second-level RSA was compatible with that observed for the box scrambling experiment, reinforcing the notion that the large-scale organization of the two pathways was reliable and stable across the two experiments. The MDS plot shows that the patterns of activation in the posterior parts of the dorsal pathway (i.e. V3a-IPS1) were highly correlated with regions of the ventral pathway (e.g., LO), while more anterior parts of the two pathways had distinctive representational structures. The statistical quantification replicated the box scrambling experiment findings, with higher correlations of the between-pathways ROIs (lateral-ventral and posterior-dorsal) compared with the correlations of the within-pathways ROIs [within ventral: F_(1,10) = 54, p=0.00002, η_p² = 0.84,, CI{0.08, 0.15}; within dorsal: F _(1,10) = 130, p=0.000001, η_p² = 0.94, CI{0.11, 0.17}]. Notably, as in the box scrambling experiment, this effect was replicated even when aIPS was excluded from the analysis, or when IPS2 was included in the posterior (rather than anterior) dorsal pathway.

Correlation between fMRI activation and object recognition performance

A key question that emerges is whether the neural patterns we have uncovered bear any relation to perceptual performance. Previous studies have observed a robust correlation between fMRI activation in object-selective regions of the ventral pathway and object recognition abilities (Avidan et al., 2002; Grill-Spector et al., 2000). Because this correlation has not been examined for the dorsal pathway in the context of non-spatial perceptual tasks (rather than, for example, working memory, Jeong and Xu, 2016), we correlated the fMRI activation for the different levels of scrambling with recognition abilities measured outside the scanner.

Box scrambling experiment

Object recognition abilities decreased as a function of scrambling, as revealed by a main effect of scrambling level in the repeated measures ANOVA [F_(1,10) = 477, p=0.000001, η_p² = 0.97], and a robust simple effect of the linear contrast [F_(1,10) = 1634, p=0.000001] (Figure 5A, black line). Given this linear relationship, and a similar linear relationship between fMRI signal and scrambling level, any correlation between the fMRI signal and behavioral performance might be attributed to the shared correlation of these variables with the level of scrambling. Therefore, we computed the partial correlations between fMRI signal and behavioral performance of each participant after regressing out the shared correlation with the level of scrambling. Consistent with previous reports, fMRI activation along the ventral pathway was correlated with perceptual performance on both the inferior and lateral surfaces (r > 0, q < 0.05; Figure 6B, black markers). The more novel finding is that dorsal-pathway activation was also correlated with perceptual behavior in multiple ROIs along the IPS (r > 0, q < 0.05; Figure 6B), suggesting a functional role for the dorsal pathway in visual perception.

Figure 6

Download asset Open asset

Figure 6—source data 1

Individual data points for Figure 6A.

https://doi.org/10.7554/eLife.27576.030

Download elife-27576-fig6-data1-v2.xlsx

Figure 6—source data 2

Individual data points for Figure 6B.

https://doi.org/10.7554/eLife.27576.031

Download elife-27576-fig6-data2-v2.xlsx

Diffeomorphic scrambling experiment

In the diffeomorphic scrambling experiment, object recognition abilities decreased as a function of scrambling, as revealed by a main effect of level of scrambling in the repeated measures ANOVA [F_(1,9)=929, p=0.0000001, η_p² = 0.99], and a robust simple effect of the linear contrast [F_(1,9)=20596, p=0.0000001] (Figure 6A, gray line). Recognition performance for the scrambled versions (levels S2, S3 and S4) observed in the diffeomorphic scrambling were lower than those obtained for the box scrambling experiment, as evident from the interaction between experiment and level of scrambling [F_(4,76)=71, p=0.0000001, η_p² = 0.78] and the planned comparisons [Fs_(1,19) > 5.7, p<0.05]. Hence, the lower fMRI sensitivity to object shape (i.e. flatter slope) in the diffeomorphic scrambling experiment could not be related to superior recognizability of the distorted images in this experiment. The comparison between the behavioral performances in the two experiments should be interpreted with caution, however, since participants in the box scrambling experiment were exposed to each object in all levels of scrambling whereas, in the diffeomorphic scrambling experiment, participants were exposed to a particular object only in one level of distortion.

As in the box scrambling experiment, the partial correlation between performance and fMRI signal, with the level of scrambling as a covariate, revealed a correlation between fMRI activation and perceptual performance on both the inferior and lateral surfaces of the ventral pathway (r > 0, q < 0.05; Figure 6B. gray markers). Importantly, in this analysis, the coupling between behavioral performance and fMRI activity was also apparent in different ROIs along the IPS (r > 0, q < 0.05.; Figure 6B. gray markers).

In light of recent findings indicating that the neural representation of object shape may be mediated not only by the ventral but also by the dorsal visual pathway, the present study was designed to characterize and compare the large-scale organization of shape processing across the two visual pathways. Three key findings were obtained. First, the two pathways followed a similar topographical organization with a posterior to anterior initial increase and then decrease in shape sensitivity. Second, the representational structures of regions in the dorsal and ventral pathways were highly correlated with each other. Third, the activation of posterior dorsal pathway regions, which were sensitive to shape, were correlated with recognition performance, suggesting that these regions might contribute to perceptual behaviors. Together, these findings challenge the binary distinction between the two pathways.

1. 1. Freud E
   2. Culham J
   3. Plaut D
   4. Behrmann M

   (2017) The large-scale organization of shape processing in the ventral and dorsal pathways

   Available at figshare under a CC0 Public Domain licence.

   https://doi.org/10.1184/R1/c.3889873.v1

1. 1. Andrew AM

   (1979) Another efficient algorithm for convex hulls in two dimensions

   Information Processing Letters 9:216–219.

   https://doi.org/10.1016/0020-0190(79)90072-3

   • Google Scholar
2. 1. Avidan G
   2. Harel M
   3. Hendler T
   4. Ben-Bashat D
   5. Zohary E
   6. Malach R

   (2002)

   Contrast sensitivity in human visual areas and its relationship to object recognition

   Journal of Neurophysiology 87:3102–3116.

   • PubMed
   • Google Scholar
3. 1. Barense MD
   2. Gaffan D
   3. Graham KS

   (2007) The human medial temporal lobe processes online representations of complex objects

   Neuropsychologia 45:2963–2974.

   https://doi.org/10.1016/j.neuropsychologia.2007.05.023

   • PubMed
   • Google Scholar
4. 1. Behrmann M
   2. Lee AC
   3. Geskin JZ
   4. Graham KS
   5. Barense MD

   (2016) Temporal lobe contribution to perceptual function: A tale of three patient groups

   Neuropsychologia 90:33–45.

   https://doi.org/10.1016/j.neuropsychologia.2016.05.002

   • PubMed
   • Google Scholar
5. 1. Benjamini Y
   2. Hochberg Y

   (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing

   Journal of the Royal Statistical Society 57:289–300.

   https://doi.org/10.2307/2346101

   • Google Scholar
6. 1. Berryhill ME
   2. Fendrich R
   3. Olson IR

   (2009) Impaired distance perception and size constancy following bilateral occipitoparietal damage

   Experimental Brain Research 194:381–393.

   https://doi.org/10.1007/s00221-009-1707-7

   • PubMed
   • Google Scholar
7. 1. Bracci S
   2. Daniels N
   3. Op de Beeck H

   (2017) Task context overrules object- and category-related representational content in the human parietal cortex

   Cerebral Cortex 27:310–321.

   https://doi.org/10.1093/cercor/bhw419

   • PubMed
   • Google Scholar
8. 1. Bracci S
   2. Op de Beeck H

   (2016) Dissociations and associations between shape and category representations in the two visual pathways

   Journal of Neuroscience 36:432–444.

   https://doi.org/10.1523/JNEUROSCI.2314-15.2016

   • PubMed
   • Google Scholar
9. 1. Brodeur MB
   2. Dionne-Dostie E
   3. Montreuil T
   4. Lepage M

   (2010) The Bank of Standardized Stimuli (BOSS), a new set of 480 normative photos of objects to be used as visual stimuli in cognitive research

   PLoS One 5:e10773.

   https://doi.org/10.1371/journal.pone.0010773

   • PubMed
   • Google Scholar
10. 1. Brodeur MB
    2. Guérard K
    3. Bouras M

    (2014) Bank of Standardized Stimuli (BOSS) phase II: 930 new normative photos

    PLoS One 9:e106953.

    https://doi.org/10.1371/journal.pone.0106953

    • PubMed
    • Google Scholar
11. 1. Bugatus L
    2. Weiner KS
    3. Grill-Spector K

    (2017) Task alters category representations in prefrontal but not high-level visual cortex

    NeuroImage 155:437–449.

    https://doi.org/10.1016/j.neuroimage.2017.03.062

    • PubMed
    • Google Scholar
12. 1. Chao LL
    2. Martin A

    (2000) Representation of manipulable man-made objects in the dorsal stream

    NeuroImage 12:478–484.

    https://doi.org/10.1006/nimg.2000.0635

    • PubMed
    • Google Scholar
13. 1. Chen J
    2. Snow JC
    3. Culham JC
    4. Goodale MA

    (2017) What role does “elongation” play in “tool-specific” activation and connectivity in the dorsal and ventral visual streams?

    Cerebral Cortex 1:1–15.

    https://doi.org/10.1093/cercor/bhx017

    • Google Scholar
14. 1. Culham JC
    2. Danckert SL
    3. DeSouza JF
    4. Gati JS
    5. Menon RS
    6. Goodale MA

    (2003) Visually guided grasping produces fMRI activation in dorsal but not ventral stream brain areas

    Experimental Brain Research 153:180–189.

    https://doi.org/10.1007/s00221-003-1591-5

    • PubMed
    • Google Scholar
15. 1. Davare M
    2. Rothwell JC
    3. Lemon RN

    (2010) Causal connectivity between the human anterior intraparietal area and premotor cortex during grasp

    Current Biology 20:176–181.

    https://doi.org/10.1016/j.cub.2009.11.063

    • PubMed
    • Google Scholar
16. 1. Denys K
    2. Vanduffel W
    3. Fize D
    4. Nelissen K
    5. Peuskens H
    6. Van Essen D
    7. Orban GA

    (2004) The processing of visual shape in the cerebral cortex of human and nonhuman primates: a functional magnetic resonance imaging study

    Journal of Neuroscience 24:2551–2565.

    https://doi.org/10.1523/JNEUROSCI.3569-03.2004

    • PubMed
    • Google Scholar
17. 1. Durand JB
    2. Nelissen K
    3. Joly O
    4. Wardak C
    5. Todd JT
    6. Norman JF
    7. Janssen P
    8. Vanduffel W
    9. Orban GA

    (2007) Anterior regions of monkey parietal cortex process visual 3D shape

    Neuron 55:493–505.

    https://doi.org/10.1016/j.neuron.2007.06.040

    • PubMed
    • Google Scholar
18. 1. Fabbri S
    2. Stubbs KM
    3. Cusack R
    4. Culham JC

    (2016) Disentangling Representations of Object and Grasp Properties in the Human Brain

    Journal of Neuroscience 36:7648–7662.

    https://doi.org/10.1523/JNEUROSCI.0313-16.2016

    • PubMed
    • Google Scholar
19. 1. Fonov V
    2. Evans AC
    3. Botteron K
    4. Almli CR
    5. McKinstry RC
    6. Collins DL
    7. Brain Development Cooperative Group

    (2011) Unbiased average age-appropriate atlases for pediatric studies

    NeuroImage 54:313–327.

    https://doi.org/10.1016/j.neuroimage.2010.07.033

    • PubMed
    • Google Scholar
20. 1. Freud E
    2. Ganel T
    3. Avidan G

    (2013) Representation of possible and impossible objects in the human visual cortex: evidence from fMRI adaptation

    NeuroImage 64:685–692.

    https://doi.org/10.1016/j.neuroimage.2012.08.070

    • PubMed
    • Google Scholar
21. 1. Freud E
    2. Ganel T
    3. Shelef I
    4. Hammer MD
    5. Avidan G
    6. Behrmann M

    (2017a) Three-dimensional representations of objects in dorsal cortex are dissociable from those in ventral cortex

    Cerebral Cortex 27:422–434.

    https://doi.org/10.1093/cercor/bhv229

    • PubMed
    • Google Scholar
22. 1. Freud E
    2. Macdonald SN
    3. Chen J
    4. Quinlan DJ
    5. Goodale MA
    6. Culham JC

    (2017b) Getting a grip on reality: Grasping movements directed to real objects and images rely on dissociable neural representations

    Cortex.

    https://doi.org/10.1016/j.cortex.2017.02.020

    • PubMed
    • Google Scholar
23. 1. Freud E
    2. Plaut DC
    3. Behrmann M

    (2016) 'What' Is Happening in the Dorsal Visual Pathway

    Trends in Cognitive Sciences 20:773–784.

    https://doi.org/10.1016/j.tics.2016.08.003

    • PubMed
    • Google Scholar
24. 1. Freud E
    2. Rosenthal G
    3. Ganel T
    4. Avidan G

    (2015) Sensitivity to object impossibility in the human visual cortex: evidence from functional connectivity

    Journal of Cognitive Neuroscience 27:1029–1043.

    https://doi.org/10.1162/jocn_a_00753

    • PubMed
    • Google Scholar
25. 1. Gallivan JP
    2. Culham JC

    (2015) Neural coding within human brain areas involved in actions

    Current Opinion in Neurobiology 33:141–149.

    https://doi.org/10.1016/j.conb.2015.03.012

    • PubMed
    • Google Scholar
26. 1. Goodale MA
    2. Milner AD
    3. Jakobson LS
    4. Carey DP

    (1991) A neurological dissociation between perceiving objects and grasping them

    Nature 349:154–156.

    https://doi.org/10.1038/349154a0

    • PubMed
    • Google Scholar
27. 1. Goodale MA
    2. Milner AD

    (1992) Separate visual pathways for perception and action

    Trends in Neurosciences 15:20–25.

    https://doi.org/10.1016/0166-2236(92)90344-8

    • PubMed
    • Google Scholar
28. 1. Greenberg AS
    2. Verstynen T
    3. Chiu YC
    4. Yantis S
    5. Schneider W
    6. Behrmann M

    (2012) Visuotopic cortical connectivity underlying attention revealed with white-matter tractography

    Journal of Neuroscience 32:2773–2782.

    https://doi.org/10.1523/JNEUROSCI.5419-11.2012

    • PubMed
    • Google Scholar
29. 1. Grill-Spector K
    2. Kushnir T
    3. Hendler T
    4. Edelman S
    5. Itzchak Y
    6. Malach R

    (1998) A sequence of object-processing stages revealed by fMRI in the human occipital lobe

    Human Brain Mapping 6:316–328.

    https://doi.org/10.1002/(SICI)1097-0193(1998)6:4<316::AID-HBM9>3.0.CO;2-6

    • PubMed
    • Google Scholar
30. 1. Grill-Spector K
    2. Kushnir T
    3. Hendler T
    4. Malach R

    (2000) The dynamics of object-selective activation correlate with recognition performance in humans

    Nature Neuroscience 3:837–843.

    https://doi.org/10.1038/77754

    • PubMed
    • Google Scholar
31. 1. Grill-Spector K
    2. Weiner KS

    (2014) The functional architecture of the ventral temporal cortex and its role in categorization

    Nature Reviews Neuroscience 15:536–548.

    https://doi.org/10.1038/nrn3747

    • PubMed
    • Google Scholar
32. 1. Hutchison RM
    2. Gallivan JP

    (2016) Functional coupling between frontoparietal and occipitotemporal pathways during action and perception

    Cortex.

    https://doi.org/10.1016/j.cortex.2016.10.020

    • PubMed
    • Google Scholar
33. 1. Janssen P
    2. Srivastava S
    3. Ombelet S
    4. Orban GA

    (2008) Coding of shape and position in macaque lateral intraparietal area

    Journal of Neuroscience 28:6679–6690.

    https://doi.org/10.1523/JNEUROSCI.0499-08.2008

    • PubMed
    • Google Scholar
34. 1. Janssen P
    2. Vogels R
    3. Orban GA

    (2000) Selectivity for 3D shape that reveals distinct areas within macaque inferior temporal cortex

    Science 288:2054–2056.

    https://doi.org/10.1126/science.288.5473.2054

    • PubMed
    • Google Scholar
35. 1. Jeong SK
    2. Xu Y

    (2016) Behaviorally Relevant Abstract Object Identity Representation in the Human Parietal Cortex

    The Journal of Neuroscience 36:1607–1619.

    https://doi.org/10.1523/JNEUROSCI.1016-15.2016

    • PubMed
    • Google Scholar
36. Book

    1. Koffka K

    (1935)

    Principles of Gestalt Pyschology

    New-York: Harcourt, Brace and World.

    • Google Scholar
37. 1. Konen CS
    2. Behrmann M
    3. Nishimura M
    4. Kastner S

    (2011) The functional neuroanatomy of object agnosia: a case study

    Neuron 71:49–60.

    https://doi.org/10.1016/j.neuron.2011.05.030

    • PubMed
    • Google Scholar
38. 1. Konen CS
    2. Kastner S

    (2008) Two hierarchically organized neural systems for object information in human visual cortex

    Nature Neuroscience 11:224–231.

    https://doi.org/10.1038/nn2036

    • PubMed
    • Google Scholar
39. 1. Kourtzi Z
    2. Kanwisher N

    (2000)

    Cortical regions involved in perceiving object shape

    Journal of Neuroscience 20:3310–3318.

    • PubMed
    • Google Scholar
40. 1. Kravitz DJ
    2. Saleem KS
    3. Baker CI
    4. Ungerleider LG
    5. Mishkin M

    (2013) The ventral visual pathway: an expanded neural framework for the processing of object quality

    Trends in Cognitive Sciences 17:26–49.

    https://doi.org/10.1016/j.tics.2012.10.011

    • PubMed
    • Google Scholar
41. 1. Kriegeskorte N
    2. Mur M
    3. Bandettini P

    (2008) Representational similarity analysis - connecting the branches of systems neuroscience

    Frontiers in Systems Neuroscience 2:4.

    https://doi.org/10.3389/neuro.06.004.2008

    • PubMed
    • Google Scholar
42. 1. Lerner Y
    2. Hendler T
    3. Ben-Bashat D
    4. Harel M
    5. Malach R

    (2001) A hierarchical axis of object processing stages in the human visual cortex

    Cerebral Cortex 11:287–297.

    https://doi.org/10.1093/cercor/11.4.287

    • PubMed
    • Google Scholar
43. 1. Lewis JW

    (2006) Cortical networks related to human use of tools

    The Neuroscientist 12:211–231.

    https://doi.org/10.1177/1073858406288327

    • PubMed
    • Google Scholar
44. 1. Malach R
    2. Reppas JB
    3. Benson RR
    4. Kwong KK
    5. Jiang H
    6. Kennedy WA
    7. Ledden PJ
    8. Brady TJ
    9. Rosen BR
    10. Tootell RB

    (1995) Object-related activity revealed by functional magnetic resonance imaging in human occipital cortex

    PNAS 92:8135–8139.

    https://doi.org/10.1073/pnas.92.18.8135

    • PubMed
    • Google Scholar
45. 1. Margalit E
    2. Shah MP
    3. Tjan BS
    4. Biederman I
    5. Keller B
    6. Brenner R

    (2016) The Lateral Occipital Complex shows no net response to object familiarity

    Journal of Vision 16:3–8.

    https://doi.org/10.1167/16.11.3

    • PubMed
    • Google Scholar
46. 1. Martin CB
    2. McLean DA
    3. O'Neil EB
    4. Köhler S

    (2013) Distinct familiarity-based response patterns for faces and buildings in perirhinal and parahippocampal cortex

    Journal of Neuroscience 33:10915–10923.

    https://doi.org/10.1523/JNEUROSCI.0126-13.2013

    • PubMed
    • Google Scholar
47. 1. Mruczek RE
    2. von Loga IS
    3. Kastner S

    (2013) The representation of tool and non-tool object information in the human intraparietal sulcus

    Journal of Neurophysiology 109:2883–2896.

    https://doi.org/10.1152/jn.00658.2012

    • PubMed
    • Google Scholar
48. 1. Murray EA
    2. Bussey TJ
    3. Saksida LM

    (2007) Visual perception and memory: a new view of medial temporal lobe function in primates and rodents

    Annual Review of Neuroscience 30:99–122.

    https://doi.org/10.1146/annurev.neuro.29.051605.113046

    • PubMed
    • Google Scholar
49. 1. Murray SO
    2. Kersten D
    3. Olshausen BA
    4. Schrater P
    5. Woods DL

    (2002) Shape perception reduces activity in human primary visual cortex

    PNAS 99:15164–15169.

    https://doi.org/10.1073/pnas.192579399

    • PubMed
    • Google Scholar
50. 1. Olman CA
    2. Davachi L
    3. Inati S

    (2009) Distortion and signal loss in medial temporal lobe

    PLoS One 4:e8160.

    https://doi.org/10.1371/journal.pone.0008160

    • PubMed
    • Google Scholar
51. 1. Op de Beeck HP
    2. Torfs K
    3. Wagemans J

    (2008) Perceived shape similarity among unfamiliar objects and the organization of the human object vision pathway

    Journal of Neuroscience 28:10111–10123.

    https://doi.org/10.1523/JNEUROSCI.2511-08.2008

    • PubMed
    • Google Scholar
52. 1. Orban GA

    (2011) The extraction of 3D shape in the visual system of human and nonhuman primates

    Annual Review of Neuroscience 34:361–388.

    https://doi.org/10.1146/annurev-neuro-061010-113819

    • PubMed
    • Google Scholar
53. Book

    1. Palmer SE

    (1999)

    Vision Science: Photons to Phenomenology

    MIT press.

    • Google Scholar
54. 1. Qiu FT
    2. von der Heydt R

    (2005) Figure and ground in the visual cortex: v2 combines stereoscopic cues with gestalt rules

    Neuron 47:155–166.

    https://doi.org/10.1016/j.neuron.2005.05.028

    • PubMed
    • Google Scholar
55. Software

    1. R Development Core Team

    (2009)

    R: A Language and Environment for Statistical Computing

    R Foundation for Statistical Computing, Vienna, Austria.
56. 1. Read SJ
    2. Vanman EJ
    3. Miller LC

    (1997) Connectionism, parallel constraint satisfaction processes, and gestalt principles: (re) introducing cognitive dynamics to social psychology

    Personality and Social Psychology Review 1:26–53.

    https://doi.org/10.1207/s15327957pspr0101_3

    • PubMed
    • Google Scholar
57. 1. Sawamura H
    2. Georgieva S
    3. Vogels R
    4. Vanduffel W
    5. Orban GA

    (2005) Using functional magnetic resonance imaging to assess adaptation and size invariance of shape processing by humans and monkeys

    Journal of Neuroscience 25:4294–4306.

    https://doi.org/10.1523/JNEUROSCI.0377-05.2005

    • PubMed
    • Google Scholar
58. 1. Stojanoski B
    2. Cusack R

    (2014) Time to wave good-bye to phase scrambling: creating controlled scrambled images using diffeomorphic transformations

    Journal of Vision 14:6–16.

    https://doi.org/10.1167/14.12.6

    • PubMed
    • Google Scholar
59. 1. Theys T
    2. Pani P
    3. van Loon J
    4. Goffin J
    5. Janssen P

    (2013) Three-dimensional shape coding in grasping circuits: a comparison between the anterior intraparietal area and ventral premotor area F5a

    Journal of Cognitive Neuroscience 25:352–364.

    https://doi.org/10.1162/jocn_a_00332

    • PubMed
    • Google Scholar
60. 1. Theys T
    2. Romero MC
    3. van Loon J
    4. Janssen P

    (2015) Shape representations in the primate dorsal visual stream

    Frontiers in Computational Neuroscience 9:43.

    https://doi.org/10.3389/fncom.2015.00043

    • PubMed
    • Google Scholar
61. Book

    1. Ungerleider LG
    2. Mishkin M

    (1982)

    Two cortical visual systems

    In: Ingle D. J, Goodale M. A, Mansfield R. J. W, editors. Analysis of Visual Behavior. Cambridge: MIT Press. pp. 549–586.

    • Google Scholar
62. 1. Valyear KF
    2. Cavina-Pratesi C
    3. Stiglick AJ
    4. Culham JC

    (2007) Does tool-related fMRI activity within the intraparietal sulcus reflect the plan to grasp?

    NeuroImage 36 Suppl 2:T94–T108.

    https://doi.org/10.1016/j.neuroimage.2007.03.031

    • PubMed
    • Google Scholar
63. 1. Van Dromme IC
    2. Premereur E
    3. Verhoef BE
    4. Vanduffel W
    5. Janssen P

    (2016) Posterior Parietal Cortex Drives Inferotemporal Activations During Three-Dimensional Object Vision

    PLOS Biology 14:e1002445.

    https://doi.org/10.1371/journal.pbio.1002445

    • PubMed
    • Google Scholar
64. 1. Visser M
    2. Jefferies E
    3. Lambon Ralph MA
    4. Ralph ML

    (2010) Semantic processing in the anterior temporal lobes: a meta-analysis of the functional neuroimaging literature

    Journal of Cognitive Neuroscience 22:1083–1094.

    https://doi.org/10.1162/jocn.2009.21309

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

    (2015) Probabilistic Maps of Visual Topography in Human Cortex

    Cerebral Cortex 25:3911–3931.

    https://doi.org/10.1093/cercor/bhu277

    • PubMed
    • Google Scholar
66. 1. Zachariou V
    2. Klatzky R
    3. Behrmann M

    (2014) Ventral and dorsal visual stream contributions to the perception of object shape and object location

    Journal of Cognitive Neuroscience 26:189–209.

    https://doi.org/10.1162/jocn_a_00475

    • PubMed
    • Google Scholar
67. 1. Zachariou V
    2. Nikas CV
    3. Safiullah ZN
    4. Gotts SJ
    5. Ungerleider LG

    (2017) Spatial mechanisms within the dorsal visual pathway contribute to the configural processing of faces

    Cerebral Cortex 27:4124–4138.

    https://doi.org/10.1093/cercor/bhw224

    • PubMed
    • Google Scholar

Author details

1. Erez Freud

   1. Department of Psychology, Carnegie Mellon University, Pittsburgh, United States
   2. Center for the Neural Basis of Cognition, Carnegie Mellon University and the University of Pittsburgh, Pittsburgh, United States

   Contribution

   Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   For correspondence

   erezfreud@gmail.com

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-3758-3855

2. Jody C Culham

   1. The Brain and Mind Institute, University of Western Ontario, London, Canada
   2. Graduate Program in Neuroscience, University of Western Ontario, London, Canada
   3. Department of Psychology, University of Western Ontario, London, Canada

   Contribution

   Conceptualization, Supervision, Funding acquisition, Investigation, Methodology, Project administration, Writing—review and editing

   Competing interests

   Reviewing editor, eLife

   "This ORCID iD identifies the author of this article:" 0000-0003-0754-2999

3. David C Plaut

   1. Department of Psychology, Carnegie Mellon University, Pittsburgh, United States
   2. Center for the Neural Basis of Cognition, Carnegie Mellon University and the University of Pittsburgh, Pittsburgh, United States

   Contribution

   Conceptualization, Supervision, Funding acquisition, Investigation, Methodology, Writing—review and editing

   For correspondence

   plaut@cmu.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-0399-7120

4. Marlene Behrmann

   1. Department of Psychology, Carnegie Mellon University, Pittsburgh, United States
   2. Center for the Neural Basis of Cognition, Carnegie Mellon University and the University of Pittsburgh, Pittsburgh, United States

   Contribution

   Conceptualization, Supervision, Funding acquisition, Investigation, Methodology, Project administration, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-3814-1015

• 3,250

  views

• 509

  downloads

• 61

  citations

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