Neuroscience Letters 567 (2014) 68–73
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Integration of visual and motor functional streams in the human brain Jorge Sepulcre a,b,∗ a Division of Nuclear Medicine and Molecular Imaging, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA b Athinioula A. Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA
h i g h l i g h t s • • • •
Streams of visual and motor cortices integrate in key multimodal brain areas. Parietal operculum-4 area serves as the major relay station for the motor stream. Direct evidences of visual and motor streams converging in mirror neuron areas. Perception, action and cognition connect in the multimodal integration network.
a r t i c l e
i n f o
Article history: Received 30 November 2013 Received in revised form 27 January 2014 Accepted 21 March 2014 Keywords: Visual streams Motor streams Visuo-motor integration Brain network Graph theory Mirror neuron system
a b s t r a c t A long-standing difficulty in brain research has been to disentangle how information flows across circuits composed by multiple local and distant cerebral areas. At the large-scale level, several brain imaging methods have contributed to the understanding of those circuits by capturing the covariance or coupling patterns of blood oxygen level-dependent (BOLD) activity between distributed brain regions. The hypothesis is that underlying information processes are closely associated to synchronized brain activity, and therefore to the functional connectivity structure of the human brain. In this study, we have used a recently developed method called stepwise functional connectivity analysis. Our results show that motor and visual connectivity merge in a multimodal integration network that links together perception, action and cognition in the human functional connectome. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction The ability to adapt to the environment through visual perception and motor responses is essential for many animals in nature, including humans. Perception and motor programs combine in our brains to produce coherent temporo-spatial representations that mediate the interaction with the external world. However, it is still under debate how from high local-modular organized areas, such as early sensory and motor cortex, the brain assembles its functional streams into multimodal and further association regions of elaboration [1,2]. In the past, neuroanatomical studies – especially in macaque monkeys – brought solid evidence about the brain transitions from primary-to-association cortex. For instance, prefrontal areas
∗ Correspondence to: Harvard University, 52 Oxford Street, Northwest Building, 280.02, Cambridge, MA 02138, USA. Tel.: +1 857 869 2078. E-mail address: [email protected] http://dx.doi.org/10.1016/j.neulet.2014.03.050 0304-3940/© 2014 Elsevier Ireland Ltd. All rights reserved.
such as BA10, BA46 and orbitofrontal cortex were proposed as final cortical destination where visual and somatomotor pathways converge [3]. Nevertheless, it has been the work of the mirror neuron system (MNS) that greatly expanded the visuo-motor integration research. The MNS is defined as the brain regions that actively engage both when individuals observe and perform the same action [4,7]. The connectivity between the ventral premotor area (vPM) at the inferior frontal gyrus and the inferior parietal lobule – often called the parieto-frontal mirror circuit – [7,8] has been proposed as the prominent system of mirror properties. Originally, the MNS was discovered during neural recordings in macaque monkey, particularly in ventral premotor area F5 and ventral inferior parietal lobule area PFG [7]. Later, brain neuroimaging studies have shown potential MNS regions in humans [7,9–18] but its significance and homology with monkey studies are still controversial [19]. For instance, the location of regions with mirror properties seems to be wider and more distributed in humans than in monkeys. Other regions such as the dorsal premotor, supplementary motor area (SMA), anterior insula (aI),
J. Sepulcre / Neuroscience Letters 567 (2014) 68–73
69
Fig. 1. (A) Illustrates the SFC method. Size of nodes represents the degree of connectivity at different link-step distances in a hypothetical network for a ROI (dark blue nodes). Node size is re-scaled in each condition for better visualization. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
medial temporal lobe, superior parietal lobule (SPL) and even primary motor areas have been described as part of the human MNS [12,15,16,20–25]. Moreover, several authors have criticized a possible over-interpretation of the MNS theory regarding its potential role in action understanding and encoding of actions [19]. The evidence that these functions are taking place in the MNS areas are sparse and divisive [19]. Other questions about brain processing of visuo-motor information remain open [26]. For instance, what is the functional brain connectivity structure that supports visuo-motor integration in humans? How does information flow from sensory to high-order cognitive areas? Where are the precise areas of convergences of the visual and motor pathways in the brain? In this study, we investigate the visual and motor functional streams across the entire human brain by using a novel neuroimaging method called stepwise functional connectivity (SFC) [2,27]. Our findings facilitate a comprehensive description of the large-scale connectivity integration of the visual and motor systems. 2. Materials and methods 2.1. Participants All analyses were based on a dataset of 100 healthy young adults (mean age = 21.3 yr, 37% male) examined before in Sepulcre et al. [27]. Therefore, the present investigation has developed from a previous study that focused only on sensory modalities of the human brain [27], extending its analysis to the motor system. Subjects were recruited as part of a neuroimaging collaborative effort across multiple laboratories at Harvard University, the Massachusetts General Hospital and the greater Boston area [28,29]. 2.2. MRI acquisition and stepwise functional connectivity analysis Scanning was acquired on a 3 T TimTrio system (Siemens, Erlangen, Germany) using the vendor 12-channel phased-array head coil. Imaging preprocessing steps were optimized for fcMRI analysis [30–32] extending an approach developed by Biswal et al. [33] (see Supplementary data for acquisition parameters and fcMRI preprocessing information).
SFC is a graph theory method [27] base on the signal couplings of spontaneous low-frequency BOLD fluctuations between brain regions – see Fig. 1 for a graphical display – also referred as intrinsic activity [33,34]. The SFC method complements the conventional intrinsic activity approach by detecting not only direct functional couplings of a brain region but also its indirect – but meaningful – associations in successive steps of connectivity. A mask of 4652 voxels covering the whole brain (cortex, subcortex, brain stem and cerebellum) was used to extract the time courses (124 time points). The FDR thresholded matrices were binarized to obtain undirected and un-weighted graphs for each individual that will serve as input data for the SFC analysis. Finally, The SFC analysis computes the degree or count of all paths that: (1) connect a given voxel in the brain to a primary cortex area of interest in (2) an exact length of connectivity distance. Connectivity distance refers here to the number of edges (link-steps) that have to be in between the voxel under analysis and the seed voxel (e.g. a voxel in the V1 visual cortex). In other words, SFC analysis finds the network pathways associated to a region of interest (ROI) without any other a priori selection constrain. Accordantly to our previous investigations, it is well-known that SFC maps reach a final stable state that collapsed into regions now considered to be the cortical hubs of the human brain [27]. The cortical hubs are the regions with the greatest number of functional connections to other areas of the brain and seem to be at the top of the brain hierarchical structure [35,36]. Only the most representative transitions of the SFC maps were displayed in the main figures (see SF 1 and 3 for further SFC distances). As a final step before cortical projection, individual SFC volumes from the SFC analysis of each subject were z-score transformed using the mean and standard deviation of the entire study sample. In order to achieve consistent maps, we only considered z-score values equivalent to a p-value lower than 0.005. Of note, the SFC is a method that detects functional connectivity patterns, but specially in the high degree range of connections and it may underestimate alternative connectivity routes that do not necessarily have numerous pathways compared to some cortical regions, such as in subcortical structures. Therefore more work is needed to fully understand the complete motor and visual functional streams, particularly the thalamus and basal ganglia participation. In this study, we used specific regions of the motor and visual cortices as regions of interest (ROIs) for the SFC analysis. A
70
J. Sepulcre / Neuroscience Letters 567 (2014) 68–73
Fig. 2. (A) Visual SFC maps in one, two, and three link-step distances. Insets are the mirror image (M) of the left hemisphere medial surface. (B) Foot-, hand- and tongue-motor SFC maps in one, two, and three link-step distances. Color scale represents the normalized SFC degree. Red color is associated with a z-score p-value lower than 0.005. Insets are magnified and thresholded versions of specific maps to aid in the visualization of results. Gray area shows the overlap map for each modality (cut-off of 1 in the normalized color scale used in the original maps). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
previous reported task activation study served to accurately select the approximate locations of the motor and somatosensory ROIs (motor foot M1/BA4, MNI: −6, −21, 73 and 10, −21, 73; motor hand M1/BA4, MNI: −30, −13, 65 and 34, −13, 65; motor tongue M1/BA4, MNI: −54, −5, 33 and 50, −5, 33); somatosensory hand (S1/BA3, MNI: −30, −29, 65); somatosensory foot (S1/BA3, MNI: −14, −37, 73) [28]. Whereas visual (V1/BA17, MNI: −6, −77, 9 and −10, −77, 9) and auditory (A1/BA22, MNI: −54, −13, 9) ROIs were selected based on Sepulcre et al. [27] and probabilistic cytoarchitectonic maps [37–39]. All ROIs had a cube dimension of eight isotropic voxels of 8 mm. Both left and right hemispheres were included in the SFC analyses, but where appropriate, only left (Fig. 2 and SF 1–4) or right (SF 4) ROIs were considered.
2.3. Visualization To aid visualization between maps, images were displayed using Caret software [40] and a normalized color scale where 0 is the intensity corresponding to z-score equal to 0 and 1.5 is the intensity corresponding to the maximum z-score value. Red color in cortical maps always represent an associated z-score p-value lower than 0.005. Fig. 2 shows only one to three link-step distances maps to highlight the findings of the study; for further distance maps see SF 1 and 3.
3. Results 3.1. Functional streams of visual and motor primary areas We investigated the functional streams of the visual primary cortex in V1 (BA 17) (Fig. 2A) and three representative motor regions – foot, hand and tongue – all corresponding to M1 primary motor cortex (BA4) (Fig. 2B). In the one and two link-step distance maps, the visual primary cortex displays strong connectivity to dorsal (letter a), lateral (letter b) and ventral (letter c) areas of the occipital lobe (Fig. 2A). In the subsequent step, all functional streams merge within a set of distributed regions located in the SPL (letter d), the parietal operculum (OP) (OP1 area; letter e), aI/vPM cortex (letter f) and dorsal anterior cingulate cortex/SMA (dACC/SMA) (Fig. 2A). The three link-step distance map consistently confirms that the main connectivity streams of primary visual cortex merge in those distributed regions (letter d , e and f ), plus, although with lesser extend, the frontal eye field (FEF) (letter g) and a dorsal intra-motor area (letter h). The stepwise connectivity maps of the motor cortex show that all three primary regions – foot, hand and tongue-related motor areas – initially display a very distinctive pattern of functional connectivity compare to the visual system. First, all motor areas connect in one link-step distance to an area in the OP known as OP4 (BA43; [41]) (letter a in foot, hand and tongue-related motor
J. Sepulcre / Neuroscience Letters 567 (2014) 68–73
71
maps in Fig. 2B). However, in subsequent steps of connectivity, the motor functional stream reach the same set of regions found in the visual system; SPL (letter b and b ), OP1 (letter c and c ), aI/vPM cortex (letter d and d ) and dACC/SMA cortex (Fig. 2B). An important point to remark is that the axis of connectivity between motor cortex and OP4 exhibits distinctive topological features comparing to previously known connectivity patterns of somatosensory and auditory streams. SF 4 shows that SFC maps of motor (foot and hand; SF 4-A), somatosensory (foot and hand; SF 4-B) and auditory (SF 4-C) primary areas all have strong direct connectivity to the OP4 region (SF 4-A, 4-B, 4-C and overlap; OP4 anatomy map from [42–44]) but they target different aspects of that area. Motor primary cortex predominantly connects to the anterior part of OP4 whereas somatosensory and auditory connect to the posterior and middle part of OP4, respectively (overlap map in SF 4). The overlap maps in Fig. 2 highlight two aspects of the previous results: (1) the distinctive connectivity maps of visual and motor modalities in the early steps of connectivity (red and blue colors) and (2) the common areas where visual and motor connectivity converge (green color). 3.2. From high modality-specific to sensory-motor integration areas The finding in the previous section that the connectivity of visual and motor areas reach the same set of distributed regions, lead to a more detailed and complementary analysis of their common convergences, as well as the contribution that each specific sensory-motor modality connectivity have in the multimodal regions. We used a combined SFC approach for these purposes (see Supplementary data). As shown in SF 2 maps, the combined SFC results confirm the common convergences of connectivity in e, f, g and h, but also highlights other areas, such as the occipitotemporal lateral – an area a little bit more anterior than area in letter b – and dorsolateral prefrontal cortex (star symbols in SF 2) that display some degree of combined visuo-motor connectivity. Moreover, this approach shows the modality-specific components of the combined SFC findings. In the one-link step distance map, regions of early connectivity present misbalanced number of functional pathways, where some modalities predominate amount others. For instance, OP4 displays a prominent motor convergence pattern (column figure a), while dorsal and ventral occipital regions have a predominant visual-related connectivity (column figures b–d). With increasing steps, the visuo-motor integration regions – also referred here as the multimodal network – emerge as the common destiny and where more balanced connectivity between modalities exist (see column figures e–e , f–f , g–g and h–h in SF 2). Interestingly, and according to previous suggestions [45], the only subcortical region that shows visuo-motor connectivity integration in our study – although with a clear dominance in intra-motor connectivity – was the subthalamic nucleus (see inset in the medial view of the one link-step distance map in SF 2; SFC degree values (mean ± SD) at MNI coordinates −14, −22, 0: visual = 0.18 ± 0.2; foot-motor = 0.22 ± 0.12; hand-motor = 0.28 ± 0.11; tongue-motor = 0.34 ± 0.22). SF 3 shows the combined SFC analysis maps until the seventh link-step distances in the left and right hemispheres.
Fig. 3. Diagram of the stepwise convergence of the visual (green nodes) and motor (red nodes) systems into the multimodal integration network (blue nodes) or mirror neuron system (MNS). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
4. Discussion
4.2. Visuo-motor connectivity integration and the mirror neuron system
uses two processing streams, one ventral, flowing through occipito-temporal areas, and another dorsal, traveling along the occipital–parietal lobes. Our findings support a complementary viewpoint. Visual cortex (green nodes in Fig. 3) streams – although divergent and parallel in early connectivity steps – converge into a common destiny in the brain network (blue nodes in Fig. 3), in regions SPL, OP1, vPM/aI and dACC/SMA. The integration of visual streams into the fronto-parietal systems seems to happen, first, through posterior axes of connectivity that may retain some parallel-like processing, but after reaching the so-called multimodal network the “parallelness” of the streams disappear (see also [27] for a detailed analysis of this point). Moreover, connectivity to FEF or dorsal premotor areas in the frontal lobe is less predominant than the connectivity to the main regions of the multimodal network. In summary, the visual streams do not seem to use parallel dorsal and ventral functional pathways to assemble with the frontal lobe but a common fronto-parietal multimodal network that already integrates ventral and dorsal communications. As for the motor system, SFC maps show that the main motor functional stream connects M1 and a secondary region in anterior OP4 (red nodes in Fig. 3) to later connect with the same network as the as the visual system does (blue nodes in Fig. 3). In this sense, our findings reflect again that the strictly modality-specific streams may be only present in the early steps of connections, while the connectivity in advanced processing regions is less modality – or even somatotopically defined. Therefore, the idea of multiple and separated specific visuo-motor pathways within the multimodal network or within fronto-parietal communications [48,49] might have to be revised or newly elaborated.
4.1. A new standpoint of the visual and motor streams In their seminal papers, Ungerleider and Mishkin [46] and Milner and Goodale [47] proposed that the visual system
By investigating successive steps of functional connections from primary areas, the present study found that motor and visual functional-related streams meet in a multimodal integration
72
J. Sepulcre / Neuroscience Letters 567 (2014) 68–73
network. Other regions such as premotor, dorso-lateral prefrontal, anterior lateral occipital or even in situ primary motor areas may be relevant for visuo-motor integration as well, although they engage fewer convergent functional pathways than the multimodal integration core (light blue nodes in Fig. 3). Overall, these findings help in understanding disperse pieces of information from previous literature in visuo-motor and MNS research by incorporating them in a network framework. We can hypothesize that if the function of a system emerges as related to its functional connections, then, it seems natural to think that those previously discovered regions supporting visuo-motor integration and MNS abilities are, in fact, multimodal convergent zones of the visual and motor streams. Our findings confirm that principle. 4.3. Linking perception–action with the internal cognition Currently, there is an intense debate about the specific functions of the MNS, particularly related to action understanding [4–6,19,50]. The discovery of the MNS have renovated the interest in motor/embodied theories of cognition [19] because many of our high-order cognitive abilities may be deep rooted in our previous conscious and unconscious experiences from sensory and motor systems. However, the fact that the multimodal network is densely stepwise-connected, on one hand, to the primary systems and, on the other hand, to the cognitive cortical hubs ([27]; see also SF 1 and 3 as a corroboration in this study), it is possible to postulate that the MNS is not acting alone in its sensory-motor processes but influenced by the hubs at the top of the brain hierarchy [35,36,51,52]. Therefore, if the function of the MNS regions is judged by its connections then the MNS must be highly modulated by both bottom-up (from sensory-motor systems) and top-down (from cortical hubs) phenomena. In other words, the MNS, and therefore the multimodal network, seems to be a privilege position for incorporating and integrating basic sensory-motor information into higher-order cognitive centers. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.neulet.2014. 03.050. References [1] M.M. Mesulam, From sensation to cognition, Brain 121 (Pt 6) (1998) 1013–1052. [2] J. Sepulcre, M.R. Sabuncu, K.A. Johnson, Network assemblies in the functional brain, Curr. Opin. Neurol. 25 (2012) 384–391. [3] E.G. Jones, T.P. Powell, An anatomical study of converging sensory pathways within the cerebral cortex of the monkey, Brain 93 (1970) 793–820. [4] G. Di Pellegrino, L. Fadiga, L. Fogassi, V. Gallese, G. Rizzolatti, Understanding motor events: a neurophysiological study, Exp. Brain Res. 91 (1992) 176–180. [5] V. Gallese, L. Fadiga, L. Fogassi, G. Rizzolatti, Action recognition in the premotor cortex, Brain 119 (Pt 2) (1996) 593–609. [6] G. Rizzolatti, L. Craighero, The mirror-neuron system, Annu. Rev. Neurosci. 27 (2004) 169–192. [7] G. Rizzolatti, C. Sinigaglia, The functional role of the parieto-frontal mirror circuit: interpretations and misinterpretations, Nat. Rev. Neurosci. 11 (2010) 264–274. [8] G. Rizzolatti, L. Fadiga, V. Gallese, L. Fogassi, Premotor cortex and the recognition of motor actions, Brain Res. Cogn. Brain Res. 3 (1996) 131–141. [9] S.T. Grafton, M.A. Arbib, L. Fadiga, G. Rizzolatti, Localization of grasp representations in humans by positron emission tomography. 2. Observation compared with imagination, Exp. Brain Res. 112 (1996) 103–111. [10] J. Decety, J. Grèzes, N. Costes, D. Perani, M. Jeannerod, E. Procyk, F. Grassi, F. Fazio, Brain activity during observation of actions. Influence of action content and subject’s strategy, Brain 120 (Pt 10) (1997) 1763–1777. [11] M. Krams, M.F. Rushworth, M.P. Deiber, R.S. Frackowiak, R.E. Passingham, The preparation, execution and suppression of copied movements in the human brain, Exp. Brain Res. 120 (1998) 386–398. [12] J. Grezes, N. Costes, J. Decety, The effects of learning and intention on the neural network involved in the perception of meaningless actions, Brain 122 (Pt 1) (1999) 1875–1887.
[13] F. Binkofski, K. Amunts, K.M. Stephan, S. Posse, T. Schormann, H.J. Freund, K. Zilles, R.J. Seitz, Broca’s region subserves imagery of motion: a combined cytoarchitectonic and fMRI study, Hum. Brain Mapp. 11 (2000) 273–285. [14] N. Nishitani, R. Hari, Temporal dynamics of cortical representation for action, Proc. Natl. Acad. Sci. U.S.A. 97 (2000) 913–918. [15] G. Buccino, F. Binkofski, G.R. Fink, L. Fadiga, L. Fogassi, V. Gallese, R.J. Seitz, K. Zilles, G. Rizzolatti, H.J. Freund, Action observation activates premotor and parietal areas in a somatotopic manner: an fMRI study, Eur. J. Neurosci. 13 (2001) 400–404. [16] L. Koski, A. Wohlschläger, H. Bekkering, R.P. Woods, M.C. Dubeau, J.C. Mazziotta, M. Iacoboni, Modulation of motor and premotor activity during imitation of target-directed actions, Cereb. Cortex 12 (2002) 847–855. [17] J.M. Kilner, A. Neal, N. Weiskopf, K.J. Friston, C.D. Frith, Evidence of mirror neurons in human inferior frontal gyrus, J. Neurosci. 29 (2009) 10153–10159. [18] N.N. Oosterhof, A.J. Wiggett, J. Diedrichsen, S.P. Tipper, P.E. Downing, Surfacebased information mapping reveals crossmodal vision–action representations in human parietal and occipitotemporal cortex, J. Neurophysiol. 104 (2010) 1077–1089. [19] G. Hickok, Eight problems for the mirror neuron theory of action understanding in monkeys and humans, J. Cogn. Neurosci. 21 (2009) 1229–1243. [20] R. Hari, N. Forss, S. Avikainen, E. Kirveskari, S. Salenius, G. Rizzolatti, Activation of human primary motor cortex during action observation: a neuromagnetic study, Proc. Natl. Acad. Sci. U.S.A. 95 (1998) 15061–15065. [21] B. Wicker, C. Keysers, J. Plailly, J.P. Royet, V. Gallese, G. Rizzolatti, Both of us disgusted in My insula: the common neural basis of seeing and feeling disgust, Neuron 40 (2003) 655–664. [22] V. Gazzola, C. Keysers, The observation and execution of actions share motor and somatosensory voxels in all tested subjects: single-subject analyses of unsmoothed fMRI data, Cereb. Cortex 19 (2009) 1239–1255. [23] S. Caspers, K. Zilles, A.R. Laird, S.B. Eickhoff, ALE meta-analysis of action observation and imitation in the human brain, Neuroimage 50 (2010) 1148–1167. [24] C. Keysers, V. Gazzola, Social neuroscience: mirror neurons recorded in humans, Curr. Biol. 20 (2010) R353–R354. [25] R. Mukamel, A.D. Ekstrom, J. Kaplan, M. Iacoboni, I. Fried, Single-neuron responses in humans during execution and observation of actions, Curr. Biol. 20 (2010) 750–756. [26] M. Glickstein, How are visual areas of the brain connected to motor areas for the sensory guidance of movement? Trends Neurosci. 23 (2000) 613–617. [27] J. Sepulcre, M.R. Sabuncu, T.B. Yeo, H. Liu, K.A. Johnson, Stepwise connectivity of the modal cortex reveals the multimodal organization of the human brain, J. Neurosci. 32 (2012) 10649–10661. [28] R.L. Buckner, F.M. Krienen, A. Castellanos, J.C. Diaz, B.T. Yeo, The organization of the human cerebellum estimated by intrinsic functional connectivity, J. Neurophysiol. 106 (2011) 2322–2345. [29] B.T. Yeo, F.M. Krienen, J. Sepulcre, M.R. Sabuncu, D. Lashkari, M. Hollinshead, J.L. Roffman, J.W. Smoller, L. Zöllei, J.R. Polimeni, B. Fischl, H. Liu, R.L. Buckner, The organization of the human cerebral cortex estimated by intrinsic functional connectivity, J. Neurophysiol. 106 (2011) 1125–1165. [30] M.D. Fox, A.Z. Snyder, J.L. Vincent, M. Corbetta, D.C. Van Essen, M.E. Raichle, The human brain is intrinsically organized into dynamic, anticorrelated functional networks, Proc. Natl. Acad. Sci. U.S.A. 102 (2005) 9673–9678. [31] J.L. Vincent, A.Z. Snyder, M.D. Fox, B.J. Shannon, J.R. Andrews, M.E. Raichle, R.L. Buckner, Coherent spontaneous activity identifies a hippocampal–parietal memory network, J. Neurophysiol. 96 (2006) 3517–3531. [32] K.R. Van Dijk, T. Hedden, A. Venkataraman, K.C. Evans, S.W. Lazar, R.L. Buckner, Intrinsic functional connectivity as a tool for human connectomics: theory, properties, and optimization, J. Neurophysiol. 103 (2010) 297–321. [33] B. Biswal, F.Z. Yetkin, V.M. Haughton, J.S. Hyde, Functional connectivity in the motor cortex of resting human brain using echo-planar MRI, Magn. Reson. Med. 34 (1995) 537–541. [34] M.D. Fox, M.E. Raichle, Spontaneous fluctuations in brain activity observed with functional magnetic resonance imaging, Nat. Rev. Neurosci. 8 (2007) 700–711. [35] R.L. Buckner, J. Sepulcre, T. Talukdar, F.M. Krienen, H. Liu, T. Hedden, J.R. Andrews-Hanna, R.A. Sperling, K.A. Johnson, Cortical hubs revealed by intrinsic functional connectivity: mapping, assessment of stability, and relation to Alzheimer’s disease, J. Neurosci. 29 (2009) 1860–1873. [36] J. Sepulcre, H. H. Liu, T. Talukdar, I. Martincorena, B.T. Thomas Yeo, R.L. Buckner, The organization of local and distant functional connectivity in the human brain, PLoS Comput. Biol. 6 (2010) e1000808. [37] K. Amunts, A. Malikovic, H. Mohlberg, T. Schormann, K. Zilles, Brodmann’s areas 17 and 18 brought into stereotaxic space-where and how variable? Neuroimage 11 (2000) 66–84. [38] S.B. Eickhoff, A. Schleicher, K. Zilles, K. Amunts, The human parietal operculum. I. Cytoarchitectonic mapping of subdivisions, Cereb. Cortex 16 (2006) 254–267. [39] S.B. Eickhoff, K. Amunts, H. Mohlberg, K. Zilles, The human parietal operculum. II. Stereotaxic maps and correlation with functional imaging results, Cereb. Cortex 16 (2006) 268–279. [40] D.C. Van Essen, D.L. Dierker, Surface-based and probabilistic atlases of primate cerebral cortex, Neuron 56 (2007) 209–225. [41] S.B. Eickhoff, C. Grefkes, K. Zilles, G.R. Fink, The somatotopic organization of cytoarchitectonic areas on the human parietal operculum, Cereb. Cortex 17 (2007) 1800–1811. [42] K. Amunts, A. Schleicher, K. Zilles, Cytoarchitecture of the cerebral cortex – more than localization, Neuroimage 37 (2007) 1061–1068.
J. Sepulcre / Neuroscience Letters 567 (2014) 68–73 [43] H. Burton, R.J. Sinclair, D.G. McLaren, Cortical network for vibrotactile attention: a fMRI study, Hum. Brain Mapp. 29 (2008) 207–221. [44] K. Zilles, K. Amunts, Centenary of Brodmann’s map – conception and fate, Nat. Rev. Neurosci. 11 (2010) 139–145. [45] M. Alegre, J. Guridi, J. Artieda, The mirror system, theory of mind and Parkinson’s disease, J. Neurol. Sci. 310 (2011) 194–196. [46] M. Mishkin, L.G. Ungerleider, Contribution of striate inputs to the visuospatial functions of parieto-preoccipital cortex in monkeys, Behav. Brain Res. 6 (1982) 57–77. [47] D. Milner, M. Goodale, The Visual Brain in Action, Oxford University Press, Oxford, UK, 1995.
73
[48] I. Faillenot, I. Toni, J. Decety, M.C. Gregoire, M. Jeannerod, Visual pathways for object-oriented action and object recognition: functional anatomy with PET, Cereb. Cortex 7 (1997) 77–85. [49] F. Binkofski, L.J. Buxbaum, Two action systems in the human brain, Brain Lang. 127 (2012) 222–229. [50] G. Hickok, C. Sinigaglia, Clarifying the role of the mirror system, Neurosci. Lett. (2012). [51] P.S. Goldman-Rakic, Topography of cognition: parallel distributed networks in primate association cortex, Annu. Rev. Neurosci. 11 (1988) 137–156. [52] S.L. Bressler, V. Menon, Large-scale brain networks in cognition: emerging methods and principles, Trends Cogn. Sci. 14 (2010) 277–290.
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Integration of visual and motor functional streams in the human brain Jorge Sepulcre a,b,∗ a Division of Nuclear Medicine and Molecular Imaging, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA b Athinioula A. Martinos Center for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA
h i g h l i g h t s • • • •
Streams of visual and motor cortices integrate in key multimodal brain areas. Parietal operculum-4 area serves as the major relay station for the motor stream. Direct evidences of visual and motor streams converging in mirror neuron areas. Perception, action and cognition connect in the multimodal integration network.
a r t i c l e
i n f o
Article history: Received 30 November 2013 Received in revised form 27 January 2014 Accepted 21 March 2014 Keywords: Visual streams Motor streams Visuo-motor integration Brain network Graph theory Mirror neuron system
a b s t r a c t A long-standing difficulty in brain research has been to disentangle how information flows across circuits composed by multiple local and distant cerebral areas. At the large-scale level, several brain imaging methods have contributed to the understanding of those circuits by capturing the covariance or coupling patterns of blood oxygen level-dependent (BOLD) activity between distributed brain regions. The hypothesis is that underlying information processes are closely associated to synchronized brain activity, and therefore to the functional connectivity structure of the human brain. In this study, we have used a recently developed method called stepwise functional connectivity analysis. Our results show that motor and visual connectivity merge in a multimodal integration network that links together perception, action and cognition in the human functional connectome. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction The ability to adapt to the environment through visual perception and motor responses is essential for many animals in nature, including humans. Perception and motor programs combine in our brains to produce coherent temporo-spatial representations that mediate the interaction with the external world. However, it is still under debate how from high local-modular organized areas, such as early sensory and motor cortex, the brain assembles its functional streams into multimodal and further association regions of elaboration [1,2]. In the past, neuroanatomical studies – especially in macaque monkeys – brought solid evidence about the brain transitions from primary-to-association cortex. For instance, prefrontal areas
∗ Correspondence to: Harvard University, 52 Oxford Street, Northwest Building, 280.02, Cambridge, MA 02138, USA. Tel.: +1 857 869 2078. E-mail address: [email protected] http://dx.doi.org/10.1016/j.neulet.2014.03.050 0304-3940/© 2014 Elsevier Ireland Ltd. All rights reserved.
such as BA10, BA46 and orbitofrontal cortex were proposed as final cortical destination where visual and somatomotor pathways converge [3]. Nevertheless, it has been the work of the mirror neuron system (MNS) that greatly expanded the visuo-motor integration research. The MNS is defined as the brain regions that actively engage both when individuals observe and perform the same action [4,7]. The connectivity between the ventral premotor area (vPM) at the inferior frontal gyrus and the inferior parietal lobule – often called the parieto-frontal mirror circuit – [7,8] has been proposed as the prominent system of mirror properties. Originally, the MNS was discovered during neural recordings in macaque monkey, particularly in ventral premotor area F5 and ventral inferior parietal lobule area PFG [7]. Later, brain neuroimaging studies have shown potential MNS regions in humans [7,9–18] but its significance and homology with monkey studies are still controversial [19]. For instance, the location of regions with mirror properties seems to be wider and more distributed in humans than in monkeys. Other regions such as the dorsal premotor, supplementary motor area (SMA), anterior insula (aI),
J. Sepulcre / Neuroscience Letters 567 (2014) 68–73
69
Fig. 1. (A) Illustrates the SFC method. Size of nodes represents the degree of connectivity at different link-step distances in a hypothetical network for a ROI (dark blue nodes). Node size is re-scaled in each condition for better visualization. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
medial temporal lobe, superior parietal lobule (SPL) and even primary motor areas have been described as part of the human MNS [12,15,16,20–25]. Moreover, several authors have criticized a possible over-interpretation of the MNS theory regarding its potential role in action understanding and encoding of actions [19]. The evidence that these functions are taking place in the MNS areas are sparse and divisive [19]. Other questions about brain processing of visuo-motor information remain open [26]. For instance, what is the functional brain connectivity structure that supports visuo-motor integration in humans? How does information flow from sensory to high-order cognitive areas? Where are the precise areas of convergences of the visual and motor pathways in the brain? In this study, we investigate the visual and motor functional streams across the entire human brain by using a novel neuroimaging method called stepwise functional connectivity (SFC) [2,27]. Our findings facilitate a comprehensive description of the large-scale connectivity integration of the visual and motor systems. 2. Materials and methods 2.1. Participants All analyses were based on a dataset of 100 healthy young adults (mean age = 21.3 yr, 37% male) examined before in Sepulcre et al. [27]. Therefore, the present investigation has developed from a previous study that focused only on sensory modalities of the human brain [27], extending its analysis to the motor system. Subjects were recruited as part of a neuroimaging collaborative effort across multiple laboratories at Harvard University, the Massachusetts General Hospital and the greater Boston area [28,29]. 2.2. MRI acquisition and stepwise functional connectivity analysis Scanning was acquired on a 3 T TimTrio system (Siemens, Erlangen, Germany) using the vendor 12-channel phased-array head coil. Imaging preprocessing steps were optimized for fcMRI analysis [30–32] extending an approach developed by Biswal et al. [33] (see Supplementary data for acquisition parameters and fcMRI preprocessing information).
SFC is a graph theory method [27] base on the signal couplings of spontaneous low-frequency BOLD fluctuations between brain regions – see Fig. 1 for a graphical display – also referred as intrinsic activity [33,34]. The SFC method complements the conventional intrinsic activity approach by detecting not only direct functional couplings of a brain region but also its indirect – but meaningful – associations in successive steps of connectivity. A mask of 4652 voxels covering the whole brain (cortex, subcortex, brain stem and cerebellum) was used to extract the time courses (124 time points). The FDR thresholded matrices were binarized to obtain undirected and un-weighted graphs for each individual that will serve as input data for the SFC analysis. Finally, The SFC analysis computes the degree or count of all paths that: (1) connect a given voxel in the brain to a primary cortex area of interest in (2) an exact length of connectivity distance. Connectivity distance refers here to the number of edges (link-steps) that have to be in between the voxel under analysis and the seed voxel (e.g. a voxel in the V1 visual cortex). In other words, SFC analysis finds the network pathways associated to a region of interest (ROI) without any other a priori selection constrain. Accordantly to our previous investigations, it is well-known that SFC maps reach a final stable state that collapsed into regions now considered to be the cortical hubs of the human brain [27]. The cortical hubs are the regions with the greatest number of functional connections to other areas of the brain and seem to be at the top of the brain hierarchical structure [35,36]. Only the most representative transitions of the SFC maps were displayed in the main figures (see SF 1 and 3 for further SFC distances). As a final step before cortical projection, individual SFC volumes from the SFC analysis of each subject were z-score transformed using the mean and standard deviation of the entire study sample. In order to achieve consistent maps, we only considered z-score values equivalent to a p-value lower than 0.005. Of note, the SFC is a method that detects functional connectivity patterns, but specially in the high degree range of connections and it may underestimate alternative connectivity routes that do not necessarily have numerous pathways compared to some cortical regions, such as in subcortical structures. Therefore more work is needed to fully understand the complete motor and visual functional streams, particularly the thalamus and basal ganglia participation. In this study, we used specific regions of the motor and visual cortices as regions of interest (ROIs) for the SFC analysis. A
70
J. Sepulcre / Neuroscience Letters 567 (2014) 68–73
Fig. 2. (A) Visual SFC maps in one, two, and three link-step distances. Insets are the mirror image (M) of the left hemisphere medial surface. (B) Foot-, hand- and tongue-motor SFC maps in one, two, and three link-step distances. Color scale represents the normalized SFC degree. Red color is associated with a z-score p-value lower than 0.005. Insets are magnified and thresholded versions of specific maps to aid in the visualization of results. Gray area shows the overlap map for each modality (cut-off of 1 in the normalized color scale used in the original maps). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
previous reported task activation study served to accurately select the approximate locations of the motor and somatosensory ROIs (motor foot M1/BA4, MNI: −6, −21, 73 and 10, −21, 73; motor hand M1/BA4, MNI: −30, −13, 65 and 34, −13, 65; motor tongue M1/BA4, MNI: −54, −5, 33 and 50, −5, 33); somatosensory hand (S1/BA3, MNI: −30, −29, 65); somatosensory foot (S1/BA3, MNI: −14, −37, 73) [28]. Whereas visual (V1/BA17, MNI: −6, −77, 9 and −10, −77, 9) and auditory (A1/BA22, MNI: −54, −13, 9) ROIs were selected based on Sepulcre et al. [27] and probabilistic cytoarchitectonic maps [37–39]. All ROIs had a cube dimension of eight isotropic voxels of 8 mm. Both left and right hemispheres were included in the SFC analyses, but where appropriate, only left (Fig. 2 and SF 1–4) or right (SF 4) ROIs were considered.
2.3. Visualization To aid visualization between maps, images were displayed using Caret software [40] and a normalized color scale where 0 is the intensity corresponding to z-score equal to 0 and 1.5 is the intensity corresponding to the maximum z-score value. Red color in cortical maps always represent an associated z-score p-value lower than 0.005. Fig. 2 shows only one to three link-step distances maps to highlight the findings of the study; for further distance maps see SF 1 and 3.
3. Results 3.1. Functional streams of visual and motor primary areas We investigated the functional streams of the visual primary cortex in V1 (BA 17) (Fig. 2A) and three representative motor regions – foot, hand and tongue – all corresponding to M1 primary motor cortex (BA4) (Fig. 2B). In the one and two link-step distance maps, the visual primary cortex displays strong connectivity to dorsal (letter a), lateral (letter b) and ventral (letter c) areas of the occipital lobe (Fig. 2A). In the subsequent step, all functional streams merge within a set of distributed regions located in the SPL (letter d), the parietal operculum (OP) (OP1 area; letter e), aI/vPM cortex (letter f) and dorsal anterior cingulate cortex/SMA (dACC/SMA) (Fig. 2A). The three link-step distance map consistently confirms that the main connectivity streams of primary visual cortex merge in those distributed regions (letter d , e and f ), plus, although with lesser extend, the frontal eye field (FEF) (letter g) and a dorsal intra-motor area (letter h). The stepwise connectivity maps of the motor cortex show that all three primary regions – foot, hand and tongue-related motor areas – initially display a very distinctive pattern of functional connectivity compare to the visual system. First, all motor areas connect in one link-step distance to an area in the OP known as OP4 (BA43; [41]) (letter a in foot, hand and tongue-related motor
J. Sepulcre / Neuroscience Letters 567 (2014) 68–73
71
maps in Fig. 2B). However, in subsequent steps of connectivity, the motor functional stream reach the same set of regions found in the visual system; SPL (letter b and b ), OP1 (letter c and c ), aI/vPM cortex (letter d and d ) and dACC/SMA cortex (Fig. 2B). An important point to remark is that the axis of connectivity between motor cortex and OP4 exhibits distinctive topological features comparing to previously known connectivity patterns of somatosensory and auditory streams. SF 4 shows that SFC maps of motor (foot and hand; SF 4-A), somatosensory (foot and hand; SF 4-B) and auditory (SF 4-C) primary areas all have strong direct connectivity to the OP4 region (SF 4-A, 4-B, 4-C and overlap; OP4 anatomy map from [42–44]) but they target different aspects of that area. Motor primary cortex predominantly connects to the anterior part of OP4 whereas somatosensory and auditory connect to the posterior and middle part of OP4, respectively (overlap map in SF 4). The overlap maps in Fig. 2 highlight two aspects of the previous results: (1) the distinctive connectivity maps of visual and motor modalities in the early steps of connectivity (red and blue colors) and (2) the common areas where visual and motor connectivity converge (green color). 3.2. From high modality-specific to sensory-motor integration areas The finding in the previous section that the connectivity of visual and motor areas reach the same set of distributed regions, lead to a more detailed and complementary analysis of their common convergences, as well as the contribution that each specific sensory-motor modality connectivity have in the multimodal regions. We used a combined SFC approach for these purposes (see Supplementary data). As shown in SF 2 maps, the combined SFC results confirm the common convergences of connectivity in e, f, g and h, but also highlights other areas, such as the occipitotemporal lateral – an area a little bit more anterior than area in letter b – and dorsolateral prefrontal cortex (star symbols in SF 2) that display some degree of combined visuo-motor connectivity. Moreover, this approach shows the modality-specific components of the combined SFC findings. In the one-link step distance map, regions of early connectivity present misbalanced number of functional pathways, where some modalities predominate amount others. For instance, OP4 displays a prominent motor convergence pattern (column figure a), while dorsal and ventral occipital regions have a predominant visual-related connectivity (column figures b–d). With increasing steps, the visuo-motor integration regions – also referred here as the multimodal network – emerge as the common destiny and where more balanced connectivity between modalities exist (see column figures e–e , f–f , g–g and h–h in SF 2). Interestingly, and according to previous suggestions [45], the only subcortical region that shows visuo-motor connectivity integration in our study – although with a clear dominance in intra-motor connectivity – was the subthalamic nucleus (see inset in the medial view of the one link-step distance map in SF 2; SFC degree values (mean ± SD) at MNI coordinates −14, −22, 0: visual = 0.18 ± 0.2; foot-motor = 0.22 ± 0.12; hand-motor = 0.28 ± 0.11; tongue-motor = 0.34 ± 0.22). SF 3 shows the combined SFC analysis maps until the seventh link-step distances in the left and right hemispheres.
Fig. 3. Diagram of the stepwise convergence of the visual (green nodes) and motor (red nodes) systems into the multimodal integration network (blue nodes) or mirror neuron system (MNS). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
4. Discussion
4.2. Visuo-motor connectivity integration and the mirror neuron system
uses two processing streams, one ventral, flowing through occipito-temporal areas, and another dorsal, traveling along the occipital–parietal lobes. Our findings support a complementary viewpoint. Visual cortex (green nodes in Fig. 3) streams – although divergent and parallel in early connectivity steps – converge into a common destiny in the brain network (blue nodes in Fig. 3), in regions SPL, OP1, vPM/aI and dACC/SMA. The integration of visual streams into the fronto-parietal systems seems to happen, first, through posterior axes of connectivity that may retain some parallel-like processing, but after reaching the so-called multimodal network the “parallelness” of the streams disappear (see also [27] for a detailed analysis of this point). Moreover, connectivity to FEF or dorsal premotor areas in the frontal lobe is less predominant than the connectivity to the main regions of the multimodal network. In summary, the visual streams do not seem to use parallel dorsal and ventral functional pathways to assemble with the frontal lobe but a common fronto-parietal multimodal network that already integrates ventral and dorsal communications. As for the motor system, SFC maps show that the main motor functional stream connects M1 and a secondary region in anterior OP4 (red nodes in Fig. 3) to later connect with the same network as the as the visual system does (blue nodes in Fig. 3). In this sense, our findings reflect again that the strictly modality-specific streams may be only present in the early steps of connections, while the connectivity in advanced processing regions is less modality – or even somatotopically defined. Therefore, the idea of multiple and separated specific visuo-motor pathways within the multimodal network or within fronto-parietal communications [48,49] might have to be revised or newly elaborated.
4.1. A new standpoint of the visual and motor streams In their seminal papers, Ungerleider and Mishkin [46] and Milner and Goodale [47] proposed that the visual system
By investigating successive steps of functional connections from primary areas, the present study found that motor and visual functional-related streams meet in a multimodal integration
72
J. Sepulcre / Neuroscience Letters 567 (2014) 68–73
network. Other regions such as premotor, dorso-lateral prefrontal, anterior lateral occipital or even in situ primary motor areas may be relevant for visuo-motor integration as well, although they engage fewer convergent functional pathways than the multimodal integration core (light blue nodes in Fig. 3). Overall, these findings help in understanding disperse pieces of information from previous literature in visuo-motor and MNS research by incorporating them in a network framework. We can hypothesize that if the function of a system emerges as related to its functional connections, then, it seems natural to think that those previously discovered regions supporting visuo-motor integration and MNS abilities are, in fact, multimodal convergent zones of the visual and motor streams. Our findings confirm that principle. 4.3. Linking perception–action with the internal cognition Currently, there is an intense debate about the specific functions of the MNS, particularly related to action understanding [4–6,19,50]. The discovery of the MNS have renovated the interest in motor/embodied theories of cognition [19] because many of our high-order cognitive abilities may be deep rooted in our previous conscious and unconscious experiences from sensory and motor systems. However, the fact that the multimodal network is densely stepwise-connected, on one hand, to the primary systems and, on the other hand, to the cognitive cortical hubs ([27]; see also SF 1 and 3 as a corroboration in this study), it is possible to postulate that the MNS is not acting alone in its sensory-motor processes but influenced by the hubs at the top of the brain hierarchy [35,36,51,52]. Therefore, if the function of the MNS regions is judged by its connections then the MNS must be highly modulated by both bottom-up (from sensory-motor systems) and top-down (from cortical hubs) phenomena. In other words, the MNS, and therefore the multimodal network, seems to be a privilege position for incorporating and integrating basic sensory-motor information into higher-order cognitive centers. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.neulet.2014. 03.050. References [1] M.M. Mesulam, From sensation to cognition, Brain 121 (Pt 6) (1998) 1013–1052. [2] J. Sepulcre, M.R. Sabuncu, K.A. Johnson, Network assemblies in the functional brain, Curr. Opin. Neurol. 25 (2012) 384–391. [3] E.G. Jones, T.P. Powell, An anatomical study of converging sensory pathways within the cerebral cortex of the monkey, Brain 93 (1970) 793–820. [4] G. Di Pellegrino, L. Fadiga, L. Fogassi, V. Gallese, G. Rizzolatti, Understanding motor events: a neurophysiological study, Exp. Brain Res. 91 (1992) 176–180. [5] V. Gallese, L. Fadiga, L. Fogassi, G. Rizzolatti, Action recognition in the premotor cortex, Brain 119 (Pt 2) (1996) 593–609. [6] G. Rizzolatti, L. Craighero, The mirror-neuron system, Annu. Rev. Neurosci. 27 (2004) 169–192. [7] G. Rizzolatti, C. Sinigaglia, The functional role of the parieto-frontal mirror circuit: interpretations and misinterpretations, Nat. Rev. Neurosci. 11 (2010) 264–274. [8] G. Rizzolatti, L. Fadiga, V. Gallese, L. Fogassi, Premotor cortex and the recognition of motor actions, Brain Res. Cogn. Brain Res. 3 (1996) 131–141. [9] S.T. Grafton, M.A. Arbib, L. Fadiga, G. Rizzolatti, Localization of grasp representations in humans by positron emission tomography. 2. Observation compared with imagination, Exp. Brain Res. 112 (1996) 103–111. [10] J. Decety, J. Grèzes, N. Costes, D. Perani, M. Jeannerod, E. Procyk, F. Grassi, F. Fazio, Brain activity during observation of actions. Influence of action content and subject’s strategy, Brain 120 (Pt 10) (1997) 1763–1777. [11] M. Krams, M.F. Rushworth, M.P. Deiber, R.S. Frackowiak, R.E. Passingham, The preparation, execution and suppression of copied movements in the human brain, Exp. Brain Res. 120 (1998) 386–398. [12] J. Grezes, N. Costes, J. Decety, The effects of learning and intention on the neural network involved in the perception of meaningless actions, Brain 122 (Pt 1) (1999) 1875–1887.
[13] F. Binkofski, K. Amunts, K.M. Stephan, S. Posse, T. Schormann, H.J. Freund, K. Zilles, R.J. Seitz, Broca’s region subserves imagery of motion: a combined cytoarchitectonic and fMRI study, Hum. Brain Mapp. 11 (2000) 273–285. [14] N. Nishitani, R. Hari, Temporal dynamics of cortical representation for action, Proc. Natl. Acad. Sci. U.S.A. 97 (2000) 913–918. [15] G. Buccino, F. Binkofski, G.R. Fink, L. Fadiga, L. Fogassi, V. Gallese, R.J. Seitz, K. Zilles, G. Rizzolatti, H.J. Freund, Action observation activates premotor and parietal areas in a somatotopic manner: an fMRI study, Eur. J. Neurosci. 13 (2001) 400–404. [16] L. Koski, A. Wohlschläger, H. Bekkering, R.P. Woods, M.C. Dubeau, J.C. Mazziotta, M. Iacoboni, Modulation of motor and premotor activity during imitation of target-directed actions, Cereb. Cortex 12 (2002) 847–855. [17] J.M. Kilner, A. Neal, N. Weiskopf, K.J. Friston, C.D. Frith, Evidence of mirror neurons in human inferior frontal gyrus, J. Neurosci. 29 (2009) 10153–10159. [18] N.N. Oosterhof, A.J. Wiggett, J. Diedrichsen, S.P. Tipper, P.E. Downing, Surfacebased information mapping reveals crossmodal vision–action representations in human parietal and occipitotemporal cortex, J. Neurophysiol. 104 (2010) 1077–1089. [19] G. Hickok, Eight problems for the mirror neuron theory of action understanding in monkeys and humans, J. Cogn. Neurosci. 21 (2009) 1229–1243. [20] R. Hari, N. Forss, S. Avikainen, E. Kirveskari, S. Salenius, G. Rizzolatti, Activation of human primary motor cortex during action observation: a neuromagnetic study, Proc. Natl. Acad. Sci. U.S.A. 95 (1998) 15061–15065. [21] B. Wicker, C. Keysers, J. Plailly, J.P. Royet, V. Gallese, G. Rizzolatti, Both of us disgusted in My insula: the common neural basis of seeing and feeling disgust, Neuron 40 (2003) 655–664. [22] V. Gazzola, C. Keysers, The observation and execution of actions share motor and somatosensory voxels in all tested subjects: single-subject analyses of unsmoothed fMRI data, Cereb. Cortex 19 (2009) 1239–1255. [23] S. Caspers, K. Zilles, A.R. Laird, S.B. Eickhoff, ALE meta-analysis of action observation and imitation in the human brain, Neuroimage 50 (2010) 1148–1167. [24] C. Keysers, V. Gazzola, Social neuroscience: mirror neurons recorded in humans, Curr. Biol. 20 (2010) R353–R354. [25] R. Mukamel, A.D. Ekstrom, J. Kaplan, M. Iacoboni, I. Fried, Single-neuron responses in humans during execution and observation of actions, Curr. Biol. 20 (2010) 750–756. [26] M. Glickstein, How are visual areas of the brain connected to motor areas for the sensory guidance of movement? Trends Neurosci. 23 (2000) 613–617. [27] J. Sepulcre, M.R. Sabuncu, T.B. Yeo, H. Liu, K.A. Johnson, Stepwise connectivity of the modal cortex reveals the multimodal organization of the human brain, J. Neurosci. 32 (2012) 10649–10661. [28] R.L. Buckner, F.M. Krienen, A. Castellanos, J.C. Diaz, B.T. Yeo, The organization of the human cerebellum estimated by intrinsic functional connectivity, J. Neurophysiol. 106 (2011) 2322–2345. [29] B.T. Yeo, F.M. Krienen, J. Sepulcre, M.R. Sabuncu, D. Lashkari, M. Hollinshead, J.L. Roffman, J.W. Smoller, L. Zöllei, J.R. Polimeni, B. Fischl, H. Liu, R.L. Buckner, The organization of the human cerebral cortex estimated by intrinsic functional connectivity, J. Neurophysiol. 106 (2011) 1125–1165. [30] M.D. Fox, A.Z. Snyder, J.L. Vincent, M. Corbetta, D.C. Van Essen, M.E. Raichle, The human brain is intrinsically organized into dynamic, anticorrelated functional networks, Proc. Natl. Acad. Sci. U.S.A. 102 (2005) 9673–9678. [31] J.L. Vincent, A.Z. Snyder, M.D. Fox, B.J. Shannon, J.R. Andrews, M.E. Raichle, R.L. Buckner, Coherent spontaneous activity identifies a hippocampal–parietal memory network, J. Neurophysiol. 96 (2006) 3517–3531. [32] K.R. Van Dijk, T. Hedden, A. Venkataraman, K.C. Evans, S.W. Lazar, R.L. Buckner, Intrinsic functional connectivity as a tool for human connectomics: theory, properties, and optimization, J. Neurophysiol. 103 (2010) 297–321. [33] B. Biswal, F.Z. Yetkin, V.M. Haughton, J.S. Hyde, Functional connectivity in the motor cortex of resting human brain using echo-planar MRI, Magn. Reson. Med. 34 (1995) 537–541. [34] M.D. Fox, M.E. Raichle, Spontaneous fluctuations in brain activity observed with functional magnetic resonance imaging, Nat. Rev. Neurosci. 8 (2007) 700–711. [35] R.L. Buckner, J. Sepulcre, T. Talukdar, F.M. Krienen, H. Liu, T. Hedden, J.R. Andrews-Hanna, R.A. Sperling, K.A. Johnson, Cortical hubs revealed by intrinsic functional connectivity: mapping, assessment of stability, and relation to Alzheimer’s disease, J. Neurosci. 29 (2009) 1860–1873. [36] J. Sepulcre, H. H. Liu, T. Talukdar, I. Martincorena, B.T. Thomas Yeo, R.L. Buckner, The organization of local and distant functional connectivity in the human brain, PLoS Comput. Biol. 6 (2010) e1000808. [37] K. Amunts, A. Malikovic, H. Mohlberg, T. Schormann, K. Zilles, Brodmann’s areas 17 and 18 brought into stereotaxic space-where and how variable? Neuroimage 11 (2000) 66–84. [38] S.B. Eickhoff, A. Schleicher, K. Zilles, K. Amunts, The human parietal operculum. I. Cytoarchitectonic mapping of subdivisions, Cereb. Cortex 16 (2006) 254–267. [39] S.B. Eickhoff, K. Amunts, H. Mohlberg, K. Zilles, The human parietal operculum. II. Stereotaxic maps and correlation with functional imaging results, Cereb. Cortex 16 (2006) 268–279. [40] D.C. Van Essen, D.L. Dierker, Surface-based and probabilistic atlases of primate cerebral cortex, Neuron 56 (2007) 209–225. [41] S.B. Eickhoff, C. Grefkes, K. Zilles, G.R. Fink, The somatotopic organization of cytoarchitectonic areas on the human parietal operculum, Cereb. Cortex 17 (2007) 1800–1811. [42] K. Amunts, A. Schleicher, K. Zilles, Cytoarchitecture of the cerebral cortex – more than localization, Neuroimage 37 (2007) 1061–1068.
J. Sepulcre / Neuroscience Letters 567 (2014) 68–73 [43] H. Burton, R.J. Sinclair, D.G. McLaren, Cortical network for vibrotactile attention: a fMRI study, Hum. Brain Mapp. 29 (2008) 207–221. [44] K. Zilles, K. Amunts, Centenary of Brodmann’s map – conception and fate, Nat. Rev. Neurosci. 11 (2010) 139–145. [45] M. Alegre, J. Guridi, J. Artieda, The mirror system, theory of mind and Parkinson’s disease, J. Neurol. Sci. 310 (2011) 194–196. [46] M. Mishkin, L.G. Ungerleider, Contribution of striate inputs to the visuospatial functions of parieto-preoccipital cortex in monkeys, Behav. Brain Res. 6 (1982) 57–77. [47] D. Milner, M. Goodale, The Visual Brain in Action, Oxford University Press, Oxford, UK, 1995.
73
[48] I. Faillenot, I. Toni, J. Decety, M.C. Gregoire, M. Jeannerod, Visual pathways for object-oriented action and object recognition: functional anatomy with PET, Cereb. Cortex 7 (1997) 77–85. [49] F. Binkofski, L.J. Buxbaum, Two action systems in the human brain, Brain Lang. 127 (2012) 222–229. [50] G. Hickok, C. Sinigaglia, Clarifying the role of the mirror system, Neurosci. Lett. (2012). [51] P.S. Goldman-Rakic, Topography of cognition: parallel distributed networks in primate association cortex, Annu. Rev. Neurosci. 11 (1988) 137–156. [52] S.L. Bressler, V. Menon, Large-scale brain networks in cognition: emerging methods and principles, Trends Cogn. Sci. 14 (2010) 277–290.
