NeuroImage 65 (2013) 488–498

Contents lists available at SciVerse ScienceDirect

NeuroImage

j ourna l homepage: www.e lsev ie r .com/ locate /yn img

Dissociation and convergence of the dorsal and ventral visual working memorystreams in the human prefrontal cortex

Emi Takahashi a,b,⁎,1, Kenichi Ohki c, Dae-Shik Kim b

a Department of Radiology, Massachusetts General Hospital, Harvard Medical School, Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA 02129, USAb Department of Anatomy and Neurobiology, Center for Biomedical Imaging, Boston University School of Medicine, Boston, MA 02118, USAc Department of Molecular Physiology, Kyushu University Graduate School of Medical Sciences, Higashi-ku, Fukuoka, Japan

⁎ Corresponding author at: Division of Newborn MedChildren's Hospital Boston, Harvard Medical School, Bos

E-mail address: emi@nmr.mgh.harvard.edu (E. Taka1 Current address: Division of Newborn Medicine, Dep

Hospital Boston, Harvard Medical School, Boston, MA 021

1053-8119/$ – see front matter © 2012 Elsevier Inc. Allhttp://dx.doi.org/10.1016/j.neuroimage.2012.10.002

a b s t r a c t

a r t i c l e i n f o

Article history:Accepted 3 October 2012Available online 12 October 2012

Keywords:Visual streamsPrefrontal cortexShort-term memoryFunctional magnetic resonance imagingDiffusion tractographyFunctional connectivity

Visual information is largely processed through two pathways in the primate brain: an object pathway fromthe primary visual cortex to the temporal cortex (ventral stream) and a spatial pathway to the parietal cortex(dorsal stream). Whether and to what extent dissociation exists in the human prefrontal cortex (PFC) haslong been debated. We examined anatomical connections from functionally defined areas in the temporaland parietal cortices to the PFC, using noninvasive functional and diffusion-weighted magnetic resonance im-aging. The right inferior frontal gyrus (IFG) received converging input from both streams, while the right su-perior frontal gyrus received input only from the dorsal stream. Interstream functional connectivity to the IFGwas dynamically recruited only when both object and spatial information were processed. These results sug-gest that the human PFC receives dissociated and converging visual pathways, and that the right IFG regionserves as an integrator of the two types of information.

© 2012 Elsevier Inc. All rights reserved.

Introduction

In the primate brain, visual information is conveyed from retinato primary visual cortex by three pathways: the magnocellular,parvocellular and koniocellular pathways (Livingstone and Hubel,1988). Although recent studies have found significant intermixing of dif-ferent pathways in the early primary visual cortex (Lachica et al., 1992;Yabuta and Callaway, 1998), and in V2 (Sincich and Horton, 2002), ithas also been suggested that such intermixing follows a specific pattern,such that new parallel streams of information are conveyed to theextrastriate cortex (Nassi and Callaway, 2009). After leaving the primaryvisual cortex, visual information is largely processed through two path-ways that involve different cortical regions: the object pathwayextending from the primary visual cortex to the temporal cortex (ventralstream) and the spatial navigation pathway to the parietal cortex (dorsalstream). Despite extensive connections between the dorsal stream andventral stream (Felleman and Van Essen, 1991), each represents differ-ent features.

The existence of these two separate visual processing pathways wasfirst proposed by Schneider (1969) followed by Ungerleider and

icine, Department of Medicine,ton, MA 02115, USA.hashi).artment of Medicine, Children's15, USA.

rights reserved.

Mishkin (1982) who, based on their lesion studies, suggested that thedorsal stream is involved in the processing of visual spatial information,such as object localization (where), and the ventral stream is involvedin the processing of visual object identification information (what)(Ungerleider and Mishkin, 1982). Since this initial proposal, it hasbeen alternatively suggested that the dorsal pathway should beknown as the ‘How’ pathway, as the visual spatial informationprocessed here provides us with information about how to interactwith objects (Goodale and Milner, 1992). For the purpose of objectrecognition, the neural focus is on the ventral stream.

This concept of separate representation of different features hasbeen further supported by lesion studies (Goodale and Milner,1992; Ungerleider and Haxby, 1994), electrophysiology studies(Gross et al., 1972; Sato et al., 1980; Bruce et al., 1981; Fuster andJervey, 1981; Desimone et al., 1984; Batuev et al., 1985; Gnadtand Andersen, 1988; Miyashita and Chang, 1988; Koch and Fuster,1989; Barash et al., 1991a,b; Miller et al., 1991; Chafee andGoldman-Rakic, 1998), and histological tracer studies (Andersenet al., 1985; Felleman and Van Essen, 1991; Kaas, 2004; Lyon, 2007)in non-human primates.

But how are object and spatial information represented in theprefrontal cortex (PFC), which is positioned at the end of the sensoryprocessing stream (Levy and Goldman-Rakic, 2000)? This questionhas long been debated in non-human primate studies of the PFC, andthere continue to be varying viewpoints about whether these types ofinformation are processed in the same or different regions of the PFC(Fuster and Alexander, 1971; Kubota and Niki, 1971; Funahashi et al.,1989; Wilson et al., 1993; O'Scalaidhe et al., 1997; Rao et al., 1997).

489E. Takahashi et al. / NeuroImage 65 (2013) 488–498

Functional magnetic resonance imaging (fMRI) studies in humanshave suggested that object (what) information is supported by the infe-rior frontal gyrus (IFG) or ventrolateral PFC (VLPFC), and that spatial(where) information is supported by the superior frontal gyrus (SFG)(Wilson et al., 1993; Courtney et al., 1996, 1998a,b; Belger et al., 1998;Mohr et al., 2006; Sala and Courtney, 2007; Volle et al., 2008). Anothermodel suggests a hierarchical organization of the PFC, with short-termmaintenance functions ascribed to the IFG/VLPFC, and higher ordernon-mnemonic functions (e.g., manipulation of items in memory) as-cribed to the dorsolateral PFC (DLPFC), which is located anterior tothe SFG (Owen et al., 1996; Petit et al., 1996; D'Esposit et al., 1998;Mohr et al., 2006). Although these models do not necessarily excludeone another, questions remain about how the posterior cortices andthe PFC interact, and how the information from the posterior corticesis represented within the PFC. The goal of the current study was to ex-plore whether there is an area in the PFC in which the informationfrom the two streams converges, and/or whether there is no such areabut connections underlying the two streams dynamically combine theinformation from the two streams.

In the macaque or rhesus monkey, the terms “VLPFC” and “DLPFC”are often used for the upper bank and lower bank of the principal sulcusrespectively, and the term “IFG” is used for an area located posterior tothe arcuate sulcus in the inferior prefrontal region. However, it is moreelusive in humans how to termmultiple areas in the PFC. In this paper,we used the term “VLPFC” in the human brain to mention a broad arealower to the inferior frontal sulcus where in part the inferior frontalgyrus is located (e.g. Gilbert and Burgess, 2008), and used the term“IFG” as a part of the VLPFC. We intended to use the term “IFG” onlyin the context of the human study and in the description of the currenthuman study (e.g. activity in the IFG).

The anatomical connections to the PFC in the non-human primatehave provided essential information about functional localization. Anumber of anatomical studies have shown that the VLPFC andDLPFC re-gions receive different connections from posterior association cortices.The VLPFC is connected to the inferotemporal cortex (Kawamura andNaito, 1984; Barbas, 1988; Barbas and Pandya, 1989), a region involvedin the representation of visual objects. In contrast, the DLPFC receivesdense projections from the parietal cortex (Mesulam et al., 1977;Petrides and Pandya, 1984; Andersen et al., 1985; Barbas andMesulam, 1985; Cavada and Goldman-Rakic, 1989), a region involvedin visuo-spatial processing. Findings showing that the DLPFC andVLPFC in non-humanprimates are the recipients of different projections(object versus spatial) have suggested a parallel organization of func-tional dissociation within the PFC, and it would therefore be reasonableto expect that regions receiving these different visual inputs are func-tionally related to the ventral and dorsal visual-processing streamswith which they are selectively connected (Goldman-Rakic, 1988).However, our understanding of human brain connections has not ad-vanced at the same level as our understanding of connections in thenon-human primate brain, because studies involving invasive tracertechniques cannot be conducted in living human subjects. Furthermore,in postmortem human brains, the current tracer techniques permit def-inition if only short-range connections.

Recent technical advances in diffusion tensor imaging (DTI) (Basseret al., 1994) have allowed us to look at long-range white matter connec-tions in the human brain (Conturo et al., 1999; Jones et al., 1999; Mori etal., 1999). The process of reconstructing the 3-dimensional pathways iscalled diffusion tractography. An increasing number of studies haveused DTI to show anatomical connections in the human brain (Conturoet al., 1999; Basser et al., 2000; Stieltjes et al., 2001; Xu et al., 2002;Behrens et al., 2003; Lehericy et al., 2004; Powell et al., 2004; for review,see Catani and Thiebaut de Schotten, 2009; Johansen-Berg andRushworth, 2009), and recent studies have used DTI to assess connectiv-ity between functionally defined regions of interest (ROIs) (Guye et al.,2003; Toosy et al., 2004; Dougherty et al., 2005; Kim et al., 2006;Takahashi et al., 2007, 2008). In this study, we use a boot-trac algorithm

(Lazar and Alexander, 2005; Takahashi et al., 2007, 2008) to create prob-abilistic maps of DTI tractography.

Given that a cortical area can be characterized by its pattern ofconnections to other areas (Schmahmann and Pandya, 2006), andthat brain functions rely on anatomical connectivity, studyingconnectivity patterns between posterior cortices and the PFC inhumans may provide an important clue to understanding theinteractions between them. Toward this end, we used diffusiontractography to examine whether anatomical connections fromthe parietal and temporal cortices project to distinct areas in thePFC, or if they converge on the same areas in the PFC. AlthoughDTI provides information about anatomical pathways in vivo,there is no functional interpretation in the reconstructed fibersthemselves. We performed a whole-brain functional connectivityanalysis, looking at the functional coupling between functionallyidentified ROIs.

Materials and methods

Subjects

We tested fifteen healthy, normally sighted subjects (7 males and8 females, aged 21–39, mean age 26), all of whom reportedthemselves to be native speakers of English, right handed, with noneurological or psychiatric history. Written informed consent, in ac-cordance with the Declaration of Helsinki, was obtained from eachsubject after the nature and possible outcomes of the study wereexplained. Image acquisition was completed at Boston UniversitySchool of Medicine, and the scanning procedures were approved bythe BU Institutional Review Board.

Stimuli

All stimuli were presented on a tangential screen positioned 1.1 mfrom the subjects. Thirty-six face stimuli from the Max Planck Insti-tute, Tübingen, Germany (http://www.kyb.mpg.de/) were used forthe ‘face’ condition. The faces were framed by ovals so that hair wasnot visible. Faces and control stimuli were grayscale on a black back-ground, and occupied a visual angle of 3.20°×3.60°; they werepresented at the center of the screen during the face run, and at oneof nine peripheral locations during the combined run, in the encodingand judgment periods (Fig. 1). Control stimuli for faces were generat-ed by randomizing phases of original face images in Fourier domains,to preserve spatial frequency components.

White dot stimuli for the ‘spatial’ condition were generated usingMatlab software (The Mathworks Inc., Natick, MA, USA). The dotstimuli were presented on a black background, at one of the 8 posi-tions located 2.50° visual angle away from the center in the encodingand judgment periods. The same white dots served as control dotstimuli, and were presented at the center of the screen.

Between the encoding and judgment periods, white crosshairsappeared at the center of the screen, while red crosshairs appearedduring inter-trial periods. All stimuli were presented using presenta-tion software from Neurobehavioral Systems Inc., Albany, CA, USA.

Procedures

The functional experiment consisted of three independent runs:‘face’, ‘spatial’, and ‘combined’ runs. fMRI data were acquired duringall runs. There were 4 blocks in each run, and one block consisted of 4trials. Each trial lasted 18 s, consisting of a 6 sec encoding periodfollowed by 6 sec maintenance, 2 sec judgment, and 4 sec inter-trialperiods (Fig. 1). In each encoding period, 3 stimuli were displayed for2 s each. Each face was presented only once throughout all of theencoding periods. Subjects were asked to keep their eyes on the centerof the screen asmuch as possible, and to remember the stimuli they had

Fig. 1. Experimental design for functional imaging.

490 E. Takahashi et al. / NeuroImage 65 (2013) 488–498

seen (faces in the ‘face’ run, locations of the dots in the ‘spatial’ run, andfaces and locations in the ‘combined’ run) for the maintenance period.In the judgment period that followed, only one stimulus was displayed,and subjects were asked to indicate by pressing one of the two buttonson a box held in the right hand whether or not the same stimulus wasshown in the encoding period of the same trial.

In the control blocks in the ‘face’ and ‘combined’ condition, noiseovals were displayed to replace the face stimuli. In the control blocks inthe ‘spatial’ condition, a white dot appeared at the center of the screen.For the control blocks, subjects pressed either one of the two buttonsrandomly after the target stimuli appeared. This task protocol has beenused in many studies to localize face and spatial representation in thebrain (Haxby et al., 1995; Courtney et al., 1996, 1998a). We used theseshort-term memory tasks to enforce stimulus encoding, and to ensurethat both the posterior cortices and PFC were recruited to complete thetasks. The ordering of the ‘face’, ‘spatial’, and ‘combined’ runs werecounterbalanced across subjects. We used face stimuli because wewanted to study object recognition that has a direct link to the real life(not abstract objects etc.). Among the objects in the real life (e.g. real ob-jects and animals), activation in the ventral pathway was most reliablyidentified by using human face stimuli. The functional experiment lastedapproximately 18 min. Eye movements were not measured. However,subjects were instructed to fixate their eye movements at the center ofthe screen during the task. Subjects reported their response by pressingone of two buttons.

Image acquisition

We used a 3 Tesla whole-body scanner (Intera, Philips) to acquiregradient-echo echo-planar (GE-EPI) T2*-weighted BOLD-sensitive im-ages and diffusion-weighted images based on spin-echo echo-planar im-aging (SE-EPI). Conventional T1-weighted structural images were alsoobtained to provide anatomical information for each subject. We usedthe following parameters for BOLD functional image acquisition: repeti-tion time (TR)=2.6 s, echo time (TE)=35 ms flip angle=90°, in-plane

matrix size 128×128, field of view (FOV)=230 mm×230 mm, voxelresolution=1.8 mm×1.8 mm, number of slices 36, slice thickness4 mm. Slice orientation was axial, and the imaging volume was alignedto cover the whole brain. Each scanning run commenced with theacquisition of 5 dummy volumes, allowing tissue magnetization toachieve a steady state, after which functional volumes were acquired(115 volumes for each run).

We used the following parameters for DWI acquisition: TR=17.1 s,TE=80 ms, in-plane matrix size 128×128, FOV 230 mm×230 mm,voxel resolution 1.8 mm×1.8 mm, number of slices=96, slicethickness=1.5 mm, with fat suppression, b=1000 s/mm2, 15 direc-tions, gradient strength=0.2 G/mm, p reduction=2.0. For each subjectwe acquired 16 data sets (15 diffusion weighted+1 non-diffusionweighted images). A total of 4 signal averages were collected to ensuresufficient signal-to-noise ratio (SNR) for high-quality tensor mapping.From these data we calculated diffusion tensors for all imaged pixels.To compensate for motion, scans were acquired separately and thenco-registered with others before averaging. Use of the sensitivityencoding (SENSE) technique reduced susceptibility artifacts significant-ly. Diffusion-weighted data were acquired in the same FOV with BOLDimages to simplify post hoc spatial registration. Subsequently, the fociof the BOLD activations were used as seeding points for diffusiontractography. The same procedures were previously applied to dissoci-ate anatomical pathways related to specific cognitive function(Takahashi et al., 2007, 2008).

Functional MRI analyses

All functional images were initially analyzed with the SPM99 soft-ware for statistical parametric mapping (Wellcome Department ofNeurology, UK), and later with custom-made Matlab programs thatwere previously used in our studies (Takahashi et al., 2007, 2008). Foreach subject,we realigned the acquired images to thefirst volume to cor-rect for headmovement. Differences in acquisition timing between eachslice were corrected by using sinc interpolation.We spatially normalized

491E. Takahashi et al. / NeuroImage 65 (2013) 488–498

each volume to a standard EPI template of 2 mm cubic voxels in theTalairach and Tournoux space (Talairach et al., 1988), using nonlinearbasis functions. Each imagewas smoothed spatially with a Gaussian ker-nel of 8 mm full-width half-maximum (FWHM), and the time serieswassmoothed temporally with a 4-sec FWHM Gaussian kernel. Slow signaldrifts were removed by high-pass filtering using cut-off periods of170 s. For each voxel, data were best-fitted (least squares) using a linearcombination of regressors. The regressors were constructed to corre-spond to each trial type for each subject, and then convolved with thestandard hemodynamic response function (HRF). We analyzed all trialswith three regressors that corresponded to encoding, maintenance, andjudgment periods, respectively. We did not separate correct and incor-rect trials within blocks. Contrasts were examined first at the single sub-ject level; then the resulting images were analyzed at the group levelusing t-tests (random effect analysis).

We set an initial height threshold of pb0.001, and then used athreshold of pb0.01 corrected for whole-brain multiple comparisonsat the cluster level, according to the SPM standard procedures (Fristonet al., 1994). The reason why we did not perform the correction fornon-sphericity of the data was based on published suggestions byFriston et al. (2005) in which they reported that the two-stage proce-dure (SPM standard procedure) with sphericity assumptions aboutthe impact of first level variance components on second-level parame-ter estimates is a very reasonable approximation in the vast majorityof experimental situations. In addition, all of the subjects who partici-pate on our study completed tasks of the same duration; hence,first-level error variance can be considered to be the same acrosssubjects.

The location of each cluster was indicated by peak voxels in the nor-malized structural images, and labeled using the Talairach andTournoux nomenclature (Talairach et al., 1988). Table 1 lists all of theactivated clusters that passed the threshold (pb0.01, corrected formul-tiple comparisons at the cluster level). If the cluster extended into mul-tiple Brodmann's areas, we separately noted the peak voxels in differentareas. Because activated regions in the FG and IPS were bilateral andcontinuous across themidline, we separated them at themidline to cre-ate regions of interest for diffusion tractography analyses.

DWI analyses

We realigned the DWI images using the diffusion toolbox in SPM2.The first images of each run were realigned to the first image of thefirst run to remove eddy current-induced distortions. All of the imageswere then averaged across the 4 runs. For each voxel, diffusion tensorand fractional anisotropy (FA) were calculated using standard

Table 1Regions activated to face stimuli (a) and to spatial stimuli (b). Only clusters with statis-tically significant activity (pb0.01) corrected for whole-brain multiple comparisons arelisted. The coordinates and their T values are at the peak voxels in each cluster, and thecoordinates and Brodmann's areas (BA) are indicated in the Talairach and Tournouxatlas space. 1 voxel corresponds to 8 mm3. See also Supplementary Table S2.

Clustersize

T-value Coordinates(x, y, z)

R/L BA

a) Face stimuli vs. controls9544 9.77 −32, −74, −6 L 37 Fusiform gyrus

34, −58, −14 R 37 Fusiform gyrus1143 6.73 18, 16, 30 R 32 Anterior cingulate cortex

6.44 28, 38, 2 R 45/46 Ventral inf. frontal gyrus317 6.28 34, −48, 52 R 40 Inferior parietal lobe688 6.20 −4, 20, 56 L 6 Medial frontal cortex294 5.73 42, 10, 28 R 44/9 Dorsal inf. frontal gyrus

b) Spatial stimuli vs. controls2077 7.96 −18, −66, 56 L 7 Intraparietal sulcus

7.86 −34, −78, 22 L 19/39 Middle temporal gyrus7.82 20, −66, 60 R 7 Intraparietal sulcus

284 5.62 34, 6, 66 R 6 Superior frontal gyrus

procedures (Mori et al., 1999). After fitting the 15 DWIs by an ellipsoidtensor we removed voxels that had very large residuals; i.e., thosevoxels with residuals exceeding 35% of the value of the apparent diffu-sion coefficients (ADC) averaged across the whole brain. In all, approx-imately 15% of voxels distributed along the edge of the brain wereremoved. We defined the starting points for diffusion tractographyusing the T maps generated by SPM99, which resulted from randomeffect analysis of 15 subjects, as described in the “Functional MRIanalyses” section. We used pb0.001 (uncorrected) as an initial heightthreshold and pb0.01 (corrected at the cluster-level) as the startingpoint criteria (Takahashi et al., 2007, 2008).

The starting points for tractographywere set at intervals of 1 mm inthe foci of fMRI activation clusters. The coordinates of the activatedclusters in the Talairach space were reverse-normalized into eachsubject's native space, and were used as the basis for fiber tracking.We then transformed the coordinates of the end points in the nativespace to the Talirach space, averaged the data across all subjects, andsuperimposed the resultingmaps onto the normalized T1-weighted im-ages in Talairach space. We did sub-voxel seeding without grid shift,and after each 0.5 mm step, an interpolated value was calculated foreach DWI. For reverse normalization and normalization of the coordi-nates we used T1 and DWI scans in the same FOV. We have previouslyused this approach to define major white matter pathways (Takahashiet al., 2007, 2008); this method provides greater accuracy than directspatial normalization of DWIs, in which resampling reduces imageresolution.

Diffusion tractography

Weused a tractography algorithmbased on themethod described inLazar et al. (2003), a tensor deflection approach, developed to overcomecrossing fiber problem (Alexander et al., 2001), that uses the entire ten-sor information instead of a single eigenvector (Lazar et al., 2003).Tractographywas not terminated in voxels where theminor eigenvaluewas minimal and therefore the reconstructed 3-dimensional tensormodel was in a “pancake” shape. When a tractography pathway en-countered such a voxel, we let tractography continue to the adjacentvoxel following the direction in the previous voxel. We have previouslyapplied this method, showing many examples of known fiber bundlesfor validation (Takahashi et al., 2007, 2008). At every position alongthe fiber trajectory a diffusion tensor was interpolated (linear) and ei-genvectors were computed. The eigenvector associated with thegreatest eigenvalue indicates the principal direction of water diffusion.Fiber tracts were propagated along this direction over a small distance(0.5 mm) to the next point, where a new diffusion tensor was interpo-lated. Tractography was terminated when the angle between two con-secutive eigenvectorswas greater than a given threshold (60°), orwhenthe FA value was smaller than a given threshold (0.14). FAb0.14–0.15has been reported to provide the best tradeoff between fewer errone-ous tracts and penetration into the white matter (Thottakara et al.,2006); we used this same criterion in our previous studies (Takahashiet al., 2007, 2008). Although the tensor deflection algorithm issuggested to overcome the crossing fiber problem, and 15 directionsthat we used are more than double of the minimum number of the dif-fusion gradient directions, having more diffusion gradient directionswould give more accurate tractography results.

Group analysis in diffusion tractography—population map

For Figs. 2A andC,weperformeddiffusion tractography for all voxelsin each activated cluster. To establish dissociation and convergence ofthe anatomical connectivity of the dorsal and ventral pathways, weused a heuristic algorithm that propagated tensorlines from both re-gions.We determined the terminal points of the tensorlines by calculat-ing across subjects the probability that an end point from a particularseed region would fall within a voxel (Takahashi et al., 2007, 2008).

Fig. 2. (Upper panel) Activation maps superimposed on a three-dimensional brain.Areas in green were active to spatial stimuli, while areas in red were active to facestimuli during the encoding phase. The image on the left side represents the righthemisphere, and includes all activated areas with a height threshold of pb0.001(uncorrected). Activated areas with a height threshold of pb0.001, corrected forwhole-brain multiple comparisons at the cluster level (pb0.01), are listed in Table 1.(Lower panel) Group analyses (n=15) of tractography from the right FG (A, B) andright IPS (C, D). Population maps (A, C) and BOOT-TRAC probabilistic maps (B, D) aresuperimposed on a T1-weighted template brain. Talairach coordinates of each sectionare shown on top of each panel. Red areas in the population maps (right FG in panel A,right IPS in panel C) represent activated areas in the right FG (A) and right IPS (C). Thedark blue arrowheads point to areas around the right SFG, while light blue arrowheadspoint to areas around the right ventral IFG. See also Supplementary Table S1 and S2.

Fig. 3. (A) Population maps of the dorsal and ventral pathways projected onto athree-dimensional template brain. Terminal points of tractography (n>5) from the acti-vated area in the right IPS are shown in green, and terminal points of tractography(n>5) from the activated area in the right FG are shown in red. Overlap areas from thetwo pathways were shown in yellow. The dark blue arrowhead indicates the right SFG,and light blue arrowheads indicate the right ventral IFG. See also Supplementary Fig. S2.(B) Activation maps superimposed on a three-dimensional brain during the encoding(dark blue) and maintenance (light blue) phases of the combined condition. The imageon the left side represents the right hemisphere, and includes all activated areas with aheight threshold of pb0.001 (uncorrected). The light blue arrowhead indicates the rightventral IFG.

492 E. Takahashi et al. / NeuroImage 65 (2013) 488–498

The resultingmapwas superimposed onto normalized T1-weighted an-atomical images. The maps, which we defined as “population maps,”were reported as the percentage of subjects in whom connectionswere found.

Estimation of tractography error by the bootstrap method—BOOT-TRACprobabilistic map

To ensure the reliability of tractography, we estimated the disper-sion errors in diffusion tractography using a statistical nonparametricbootstrap method (BOOT-TRAC) (Lazar and Alexander, 2005) to eval-uate intra- and inter-subject variability (Figs. 2B and D). For each gra-dient direction, we randomly selected and averaged two of fourdatasets (volumes), and obtained an averaged 15-direction dataset,diffusion tensors, and reconstructed fibers. To obtain a BOOT-TRAC

probabilistic map from multiple seeds we first performed fiber track-ing from all seed points, for each bootstrap sample. We repeated thisprocedure 100 times and created probabilistic maps based on howmany times, out of 100, fibers passed each voxel in the brain. Weassigned a value of “1” to voxels that were passed by at one or morefibers, and “0” to voxels not passed by any fiber. We repeated thisprocedures for 100 bootstrap samples and determined probabilitiesfor each “1” voxel; we refer to the resulting maps as “BOOT-TRACprobabilistic maps” (Takahashi et al., 2007, 2008).

Estimation of anatomical convergence

For display purposes, we projected the population maps onto a3-dimensional template brain (Fig. 3A). This procedure was done bymodifying one of the programs in the SPM package that is used to pro-ject activated regions onto the 3-dimensional template brain. Instead ofactivation maps, we used the population maps after thresholding themat 5. The threshold of 5 was set for the population analyses only for a vi-sualization purpose. We displayed the terminal points of tractographyfrom the activated areas in the right IPS (terminal areas are shown ingreen) and in the right FG (terminal areas shown in red). Overlapareas are shown in yellow.

To further estimate anatomical convergence in a quantitative way,we examinedwhether the anatomical pathways from the dorsal stream(right IPS) and ventral stream (right FG) converged on the same voxels.We noted areas that were continuous across more than 100 voxels, andthat showedmore than 50% BOOT-TRAC probabilities in both pathwaysfrom the right IPS and the right FG (Supplementary Table S1). Only peakvoxels located more than 16 mm apart are listed.

Functional connectivity analyses

We performed functional connectivity analysis of the activatedregions (McIntosh et al., 1994) using the same procedure we used inour previous study (Takahashi et al., 2008). The fMRI signals werepreprocessed for realignment, slice timing correction, normalizationand spatial smoothing, in the same way as described above (fMRIanalyses section). We then removed low-frequency drifts in the timecourse in each voxel. We did not include the realignment parameters in

493E. Takahashi et al. / NeuroImage 65 (2013) 488–498

the design matrix for the functional connectivity analyses. Power et al.(2012) reported that many long-distance correlations are decreased bysubject motion, whereas many short-distance correlations are increased.In the current study,we focused on long-distance anatomical connectionsand functional correlation, so the subject motion could decrease the cor-relations but we may not falsely detect the long-distance correlations.

Given the spatial smoothing of the fMRI data, the activity of a singlevoxel can also be considered representative of the activity of the regionaround the voxel. Peak voxelswere used for functional connectivity anal-ysis (Talairach coordinates 34,−58,−14 in the right FG and 20,−66, 60in the right IPS). We obtained maps of the correlation coefficients be-tween time courses in the peak voxel (either the right FG or right IPS)and calculated other voxels for individual subjects. To make inferencesabout the functional connectivity encoded by these correlation coeffi-cients at the between-subject level, we transformed them into summarystatistics using Fisher's Z Transform. The resulting within-subject mapswere then passed to a second-level or between-subject one-samplet-test to generate SPMs, using standardmethods. T valueswere correctedfor multiple comparisons for a whole brain at the cluster level (pb0.01),using pb0.001 (uncorrected) as an initial height threshold, to be compa-rable to the fMRI analyses.

Results

Subject performance

Reaction times were 2250±158 ms for the face task, 2120±183 ms for the spatial task, and 2428±384 ms for the combined task.The percentages of correct response in the total judgments were 100%in the face and spatial tasks, and 86.0±4.6% for the combined task.

Dissociation of activation to face and spatial stimuli

In response to face presentation, blood-oxygen level-dependent(BOLD) activity was observed in the bilateral fusiform gyri (FG;Brodmann's Area [BA] 37), and in two regions in the right inferior frontalgyrus (IFG; BA 45/46 (dorsal IFG) and BA 44/9 (ventral IFG)) (Fig. 2,upper panel, Table 1a; n=15; pb0.01, corrected for whole-brain multi-ple comparisons at cluster level, according to the SPM standard proce-dures (Friston et al., 1994)). We note that although we use IFG in thisResults section, where we discuss only human data, when comparingacross species in the Introduction and Discussion sections, we used IFG/VLPFC in reference to humans and VLPFC in reference to non-human pri-mates.We also observed activation in the right anterior cingulate cortices(ACC; BA 32), a region in the left medial frontal cortex (MFC; BA 6), theright MFC (BA 9), and the right inferior parietal lobe (IPL; BA 40).

In contrast, activation in response to spatial stimuli was observed inthe bilateral intraparietal sulcus (IPS; BA 7) extending to left visualareas (BA 19/39), and in the right SFG (BA 6) (Fig. 2, upper panel,Table 1b). The activated region in the right SFG (BA 6; Talairach coordi-nate 34, 6, 66) was located more dorsal from the frontal eye field (FEF;BA 8), which suggested that SFG activity was not due to subjects' eyemovements.We notedweak activation in the left SFG region in responseto the spatial stimuli, below the threshold (p>0.01 corrected for multi-ple comparisons). Significant functional activation in response to faceand spatial stimuli was clearly dissociated in both hemispheres. Theright IPS region activated by spatial stimuli was located dorsal to theright IPL region activated by the face stimuli, and there was no overlap(at threshold of pb0.01 corrected for whole-brain multiple comparisonsat the cluster level).

Anatomical connectivity from the posterior cortices to the PFC

We next examined how the dorsal (parietal cortex) and ventral(temporal cortex) streams anatomically project to the PFC. Weperformed diffusion tractography to explore anatomical connections

from the activated regions. Supplementary Fig. S1 shows all of thetractography fibers in one subject, from the right FG activated by facestimuli (red) and from the right IPS activated by spatial stimuli(green). The tractography pathways from the right FG were restrictedprimarily to the ventral areas, projecting to the right temporal cortexand the right IFG. In contrast, the right IPS activated by spatial stimuliprojected to both dorsal and ventral areas: the right SFG, the right tem-poral cortex, and the right IFG.

The lower panels of Figs. 2(A and C) show the percentage of subjectsinwhomwe found connections from the right FG (Fig. 2A) and the rightIPS (Fig. 2C) to each voxel (“population map”; see Materials andmethods section for more details). We found a high probability of con-nections from the right FG in the right temporal pole and in the rightIFG; similarly, we found high probability of connections from the rightIPS in the right SFG and IFG. Figs. 2B and D show the result of theBOOT-TRAC averaged across all subjects. These BOOT-TRAC probabilis-tic maps evaluate both intra- and inter-subject variability on finalreconstructed fibers. The results were very similar to what we observedin the “populationmap” (Figs. 2A and C), confirming that our results arereliable even when accounting for dispersion error.

The populationmaps of tractography from the right FG (Fig. 2A) andthe right IPS (Fig. 2C)were projected on a 3-dimensional template brain(Fig. 3A). As described above, the ventral pathways (Fig. 3A red)remained mainly in the ventral areas, and the dorsal pathways(Fig. 3A green) projected to the right SFG (an arrowhead in dark blue)as well as to the right IFG (arrowheads in light blue). Overlap areas(Fig. 3A yellow) were located in areas from the right posteriorparieto-occipital junction to the right middle temporal gyrus, the rightparahippocampal gyrus and hippocampus, and the right IFG. The rightIFG was the only PFC areawhere the two streams showed convergence.In the left hemisphere, pathways from the left FG and left IPS projectedto similar areas in the right hemisphere, but there was no convergingarea in the left PFC (Fig. 3A, see also Supplementary Fig. S2).

Intra- and inter-stream anatomical connectivity

To address connectivity within streams (dorsal to dorsal or ventralto ventral) and across streams (dorsal to ventral or ventral to dorsal),we examined connections between activated areas, across all sub-jects. We first focused on five areas: right IPS and right SFG regionsactivated by spatial stimuli (dorsal stream), and right FG and tworight IFG regions activated by face stimuli (ventral stream). Werefer to the two right IFG regions activated by face stimuli as rightdorsal IFG (BA 44/9) and right ventral IFG (BA 45/46). Intra-streamanatomical connections within the ventral stream were found be-tween the right FG and right ventral IFG in 14 subjects (93.3%), andbetween the right FG and right dorsal IFG in 10 subjects (66.7%).Intra-stream connections within the dorsal stream were found be-tween the right IPS and the right SFG in 12 subjects (80.0%). Overall,we found very consistent intra-stream connections with high proba-bility, which likely support the functional dissociation in the PFC dur-ing the timeframe of our study. Interestingly, we also found a highprobability of some inter-stream connections. The right IPS was ana-tomically connected to the right dorsal IFG in 12 subjects (80.0%), andto the right ventral IFG in 12 subjects (80.0%). In contrast, the right FGwas connected to the right SFG in only 1 subject (6.70%). We alsofound inter-stream connections between posterior areas: the rightFG and right IPS were connected in 15 subjects (100%). Connectivityamong all activated areas is reported in Supplementary Table S2.

Examination of voxel-level convergence of anatomical connectivity

We examined for projection overlaps from the dorsal stream(right IPS) and ventral stream (right FG). In the PFC, we found thatprojection overlap was highly probable (more than 50%) only in theright ventral IFG (BA 47/10, Talairach coordinate 48, 48, −8), even

Table 3Magnitude of functional connectivity along inter-stream connections observed fromdiffusion tractography. The asterisk shows significant connectivity. FG: fusiformgyrus, vIFG: ventral inferior frontal gyrus, IPS: intraparietal sulcus, SFG: superior frontalgyrus.

Areas Taskcondition

Average Z value(n=15)

Significance(n=15)

Inter-stream connectivityRight IPS and right vIFG Spatial 0.146 p=0.020

Face 0.170 p=0.021Combine 0.203 pb0.001*

Right FG and right SFG Spatial 0.138 p=0.015Face 0.115 p=0.017Combine 0.084 p=0.014

494 E. Takahashi et al. / NeuroImage 65 (2013) 488–498

though we found overlapping projections in other posterior areas(Supplementary Table S1a). In the left hemisphere, we saw no con-vergence area in the left PFC (Supplementary Table S1b), consistentwith the results shown in the previous section (Fig. 3A and Supple-mentary Fig. S2).

As described above, the right dorsal IFG region activated by facestimuli received anatomical connections from both the FG and IPS(Supplementary Table S2), but was not found to have overlappingprojections at the voxel level from the FG and IPS (SupplementaryTable S1). This was because projections from the IPS and FG terminat-ed at different places in these areas.

Intra- and inter-stream functional connectivity

We found that two regions in the right IFG were connected to boththe dorsal stream (IPS) and ventral stream (FG). The right SFG acti-vated by spatial stimuli was connected to the right IPS only, not tothe right FG. The inter-stream anatomical connectivity between theright IPS (dorsal) and the right IFG (ventral) seemed to disagreewith the functional dissociation in the PFC that we observed in thisstudy, although it may be in line with previous observations of IFG ac-tivation in response to both spatial and object short-term memorytasks (Smith et al., 1995; Owen, 1996; D'Esposit et al., 1997; Raoet al., 1997; Nystrom et al., 2000; Postle et al., 2000).

To evaluate the functional strength of connectivity between theseareas, we examined functional correlation among these areas duringboth the face task and the spatial task. In agreement with anatomicalconnectivity results, intra-stream functional connectivity from theright FG to the right ventral IFG, and from the right IPS to the rightSFG was found during both face and spatial tasks (Table 2). We foundthat inter-stream functional connectivity between the right IPS andthe right ventral IFG was not significant in either the object-only orthe spatial-only conditions (Table 3).

Interstream functional connectivity during integration of two streams

Here, we hypothesized that the interstream connectivity betweenthe right IPS and the right ventral IFG is important onlywhen combinedobject and spatial information is processed in the brain. To test this hy-pothesis, we asked subjects to remember both the positions and identi-ties of faces in combination (seeMaterials andmethods).We found thatthe right IFG region was specifically active during the maintenancephase, not during the perception or retrieval phases of the combinedcondition (Fig. 3B). More interestingly, we found that functional con-nectivity along this interstream pathway became significant duringthe combined condition (Table 3). These results suggest that theinterstream pathway is dynamically recruited only when combined ob-ject and spatial information is processed in the brain.

Discussion

In this study, two parallel pathways between the posterior corticesto the PFC were observed from the right FG to the right ventral IFG/

Table 2Magnitude of functional connectivity along intra-stream connections observed fromdiffusion tractography. The asterisks show significant connectivity. FG: fusiformgyrus, vIFG: ventral inferior frontal gyrus, IPS: intraparietal sulcus, SFG: superior fron-tal gyrus.

Areas Task condition Average Z value(n=15)

Significance(n=15)

Intra-stream connectivityRight FG and right vIFG Spatial 0.299 pb0.001*

Face 0.214 pb0.001*Right IPS and right SFG Spatial 0.508 pb0.001*

Face 0.289 pb0.001*

VLPFC, and from the right IPS to the right SFG. The dorsal pathwaywas activated by spatial information,while the ventral pathwaywas ac-tivated by object (face) information. The right SFG was connected onlyto the right IPS, not to the right FG. This suggested that the right SFGreceives mainly spatial information from the right IPS. We also identi-fied pathways from the right IPS to the right ventral IFG/VLPFC acrossthe dorsal and ventral pathways. The right ventral IFG/VLPFC regionreceived anatomical connections from both the right FG and rightIPS. Interestingly, we found that functional connectivity along thisinterstream pathway became significant during the combined condi-tion. These results suggest that the right ventral IFG region serves asan integrator of the two types of information during short-termmemo-ry. In summary, our results suggest that the right ventral IFG is one pre-frontal area in the human brain to receive converging long-rangeanatomical pathways from both the dorsal and ventral visual streams,and the right ventral IFG region may incorporate spatial informationfrom the right IPS with object information from the right FG, so as tocombine and maintain both types of information.

Intra-stream long-range anatomical connectivity—comparisons withnon-human primate studies

Although anatomical connections in non-human primates havebeen well characterized between the posterior cortices and the PFC, ev-idence for connectivity in humans is still limited, in part by technologi-cal limits. Our tractography results that showed pathways running fromthe right IPS to the right SFG likely correspond to the human homolog ofthe second branch of the superior longitudinal fasciculus (SLF II), as de-scribed in the human brain (Thiebaut de Schotten et al., 2005) andmon-key brain (Schmahmann, and Pandya, 2006). In the monkey, the SLF IIoriginates in the caudal inferior parietal lobe (corresponding to thehuman angular gyrus) and the occipito-parietal area, and projects tothe dorsolateral prefrontal cortex (Schmahmann, and Pandya, 2006);these results mirror our tractography results showing connections be-tween the right IPS and the right SFG, originating from broad areasaround the parieto-occipital junction that includes the ventral end ofthe right IPS and dorsal edge of the right FG.

Our results on the pathway from the right FG to the right ventral IFG/VLPFC also showed strong agreementwith previous observations in theventral course of the inferior fronto-occipital fasciculus (IFOF) in thehuman brain, which connects the occipital/posterior temporal lobesand the inferior frontal lobe (Mori, 2007; Catani and Thiebaut deSchotten, 2009). In non-human primates, the uncinate fasciculus isthe major connection from the temporal lobe to the frontal lobe,while in humans, diffusion tractography studies have repeatedly re-vealed that the ventral pathway of the IFOF originates from the occipitaland posterior temporal lobes and courses directly to the ventral frontallobe (Mori, 2007; Takahashi et al., 2007), and that the uncinate fascicu-lus originates from the anterior temporal lobe and courses to the ventralfrontal lobe (Mori, 2007). The ventral pathways of the IFOF seem to beunique to the human brain; previous studies have led researchers to

495E. Takahashi et al. / NeuroImage 65 (2013) 488–498

suspect that the IFOF is related to visual processing (Catani andThiebaut de Schotten, 2009), and that its disconnection differentiatesthe ventral frontal cortex frommore posterior sources of visual input re-lated to object identification (Urbanski et al., 2008).

Inter-stream long-range anatomical connectivity—comparisons withnon-human primate studies

Interestingly, we identified inter-stream anatomical pathways fromthe right IPS to the right ventral IFG/VLPFC region. Based on the avail-able literature, this pathway may correspond to a branch of the IFOFin humans that courses from the parietal lobe to the frontal lobe(Mori, 2007). In non-human primates, the DLPFC receives dense con-nections particularly from the IPS and the superior parietal cortex, lessso from the inferior parietal cortex. The IFG that is located posterior tothe VLPFC (posterior to the arcuate sulcus), is connected not only tothe infero-temporal cortex, but also to the inferior parietal cortex(Schmahmann and Pandya, 2006). The third branch of the SLF (SLF III)originates from the posterior parieto-occipital junction and surround-ing areas, and courses to the VLPFC; tracer studies (Schmahmann andPandya, 2006) and tractography studies (Schmahmann et al., 2007)demonstrated that the monkey SLF III pathway may be related to ourtractography results showing connections from the right IPS to theright ventral IFG. However, it is still not clearwhether the SLF III inmon-keys correlates to a branch of the IFOF in humans. The anatomical con-verging region that we observed in the ventral IFG/VLPFC (BA 47/10 inhumans) seems to be located more anterior ventral than the terminalareas of the SLF III in monkeys (ventral BA 6, ventral BA 9/46, and BA44 in monkeys), and more anterior ventral than suggested possiblehuman frontal areas that receive pathways from the inferior parietallobe (Petrides et al., 2012). Further comparative examinations betweenhumans and monkeys will be necessary in order to conclude whethermonkeys have an anatomical converging area in the PFC, and if theydo, how that area correlates to the ventral IFG/VLPFC area in humans.

Role of the right ventral IFG/VLPFC

The right ventral IFG/VLPFC in humans may represent a convergencezone of the two streams of visual processing: (1) the occipito-temporo-frontal stream, which is dedicated to object processing andpasses through the IFOF and the uncinate fasciculus, and (2) theparieto-frontal network, which is presumably connected by the humanhomolog of the third branch of the SLF. We speculate that this areamay play an important role in integrating object and spatial information.Indeed, in monkeys, neurons in the lateral PFC respond to both the loca-tion and identity of previously presented visual objects, which allow theintegration of “what” and “where” information (Rao et al., 1997). IFG/VLPFC is, however, strongly involved in processing of working memory,and we are not able to completely exclude the possibility that the con-nectivity between the IPS and IFG was recruited because our study in-volved working memory tasks. Given that the combination of multiplefeatures is an important aspect of working memory processing, it isdifficult to dissociate those features—this remains a challenge for futurestudies. Further correlation studies for functional and anatomical con-nectivity in humans and monkeys will be important to clarify the roleof the IFG/VLPFC in integrating the information from the posterior corti-ces. A possible way to separate these two processes is to use a workingmemory task that needs the combined object and spatial informationfor completing the task, and to compare activated regions by anotherworking memory task that can be completed without combining thetwo types of information but by other strategies.

We observed activation in response to face stimuli in two right IFG/VLPFC regions: the right dorsal IFG/VLPFC (BA 44/9) and the right ven-tral IFG/VLPFC (BA 45/47); the results are consistent with previous re-ports (Haxby et al., 1995; Fiez et al., 1996; McCarthy et al., 1996;Courtney et al., 1996, 1997; Cohen et al., 1997) that referred to these

regions as the posterior mid-frontal cortex and inferior frontal cortex.These studies also found activation in the anterior mid-frontal cortex(BA 46)/DLPFC, which was more active during a memory delay period(Courtney et al., 1997). We did not find this activation, probablybecause our study included a relatively small number of items (3 loca-tions) and short delay (6 s). Courtney and colleagues reported thatsuch a short delay period with low memory load results in righthemisphere-dominant activation (Courtney et al., 1996), which isconsistent with our results. We also found right hemisphere-dominantactivation in response to spatial stimuli in a right SFG region, as has alsobeen seen in previous reports (Mellet et al., 1996; Petit et al., 1996;Courtney et al., 1996, 1998a; Mohr et al., 2006).

Hemispheric asymmetry of brain function and anatomical connectivity

In this study, we focused mainly on activation and connectivity inthe right hemisphere, because activation to face or spatial stimuli occurspredominantly in the right hemisphere. In the left hemisphere, wefound somewhat different results. While we found no anatomical con-vergence area in the left PFC, the right ventral IFG/VLPFC was a conver-gence area. Several factors may explain this hemispheric asymmetry.One possibility is that the observed hemispheric asymmetry wasinfluenced by handedness. For example, recent imaging studies havereported the relationships between handedness and brain activity dur-ing motor, memory, and language tasks (e.g. Cuzzocreo et al., 2009;Propper et al., 2010; for review, Hatta, 2007) and some of themsuggested that handedness reflects the laterality of hemispheric brainactivity in task-related brain regions. However, some other studies didnot show such clear relationships (Hatta, 2007), and further examina-tion should be done to clarify the relationships between the degree ofhandedness and deviation of functional organization in each cognitiveevent.

The other possibilities could be related to our way of data process-ing. First, the sizes of the activated areas in the IPS and FG were largerin the right hemisphere than in the left hemisphere. Since we exploredpathways from regions of interests (activated areas), the observedlaterality of the pathways could be largely due to the locations andsizes of the activated areas. Larger activated areas in the right hemi-sphere are likely to be related to larger numbers of connections fromthe activated regions, which could result in a more comprehensive net-work. Second, the sizes of the activated areas depend on the thresholdused in this study. It is possible that outer regions in the activatedarea in the left IPS might be weakly activated at a sub-threshold.These outer regions might have anatomical connections with the leftventral IFG/VLPFC. Third, anatomical connection patterns themselvesmay have hemispheric asymmetry. Given that functional hemisphericasymmetry has been repeatedly found in the human brain, it is not sur-prising to observe such asymmetry in anatomical connections inhumans. However, it may not be very likely that the major anatomicalbundles would have such large hemispheric asymmetry to the extentwe observed between the IPS and the ventral IFG/VLPFC in this study.

Activation in the right parietal lobe

In the right parietal lobe, we found activation to spatial stimuli in theright IPS, and activation to face stimuli in the right IPL although their ac-tivated regions did not overlap at the threshold of pb0.01 (corrected).This does not necessarily go against the idea of representation of spatialinformation in the parietal cortex. The right IPL region may processsome kind of spatial properties of face stimuli, or may process detailedfeatures of the face stimuli themselves. Our experimental design doesnot explicitly distinguish these two possibilities, but if the right IPL pro-cesses spatial properties of the face stimuli, spatial information togetherwith the information in the right IPSmay be transferred via the IPS-IFG/VLPFC pathway. If the right IPL itself processes face information, the twomodalities of information could converge to some extent in the local

496 E. Takahashi et al. / NeuroImage 65 (2013) 488–498

connections within the right parietal lobe before they are furtherprocessed in the PFC (Konen and Kastner, 2008).

The hierarchical model represents non-mnemonic higher orderfunctions in the PFC, such that manipulation of items in memory is as-cribed to the anterior mid-frontal cortex (BA46), and short-termmain-tenance functions are ascribed to the IFG (BA 45/47) (Owen et al., 1996;Petit et al., 1996; D'Esposit et al., 1998; Mohr et al., 2006). In our previ-ous study, we found anatomical connectivity between these areas, aswell as two parallel pathways between these two PFC areas and theFG (Takahashi et al., 2007). Although we did not find activation in theanterior mid-frontal cortex/DLPFC in this study, these connectionswould be important for manipulating information maintained in theIFG/VLPFC (Petit et al., 1996).

Our findings on the existence of the inter-stream anatomical connec-tivity between the right IPS and right ventral IFG/VLPFC may be in linewith previous observations that both spatial and object short-termmemory tasks activated the IFG/VLPFC (e.g., Owen, 1997; D'Esposit etal., 1998). In this study, however, we did not find activation to spatialstimuli in the right IFG/VLPFC, probably because of our relatively smallnumber of items (3 locations) and short delay (6 s). Indeed, in previousstudies, direct comparison of activity during face and spatial stimuli re-vealed functional dissociation between ventral and dorsal frontal corti-ces; with face information processed in the IFG/VLPFC and spatialinformation in the SFG (e.g., Courtney et al., 1998a). Thus, the functionaldissociation of the dorsal and ventral frontal areasmight not be absolute,but relative, particularly in the ventral regions.

Functional connectivity along anatomical connectivity

In agreement with the functional dissociation, we found that theinter-stream functional connectivity along this pathwaywasweak dur-ing the object-only and spatial-only conditions. This result suggests thatanatomical inter-stream connectivity exists between the right IPS andright ventral IFG/VLPFC, but its functional strength is weak duringtasks that require only one type of information, that is, either face orspatial information. This may explain why the ventral IFG/VLPFC re-gions were preferentially activated by face stimuli, but were also some-times activated by spatial stimuli. Thus, the functional dissociation inthe PFC may be mutually supported by anatomical connectivity andfunctional connectivity. Interestingly, we found that functional connec-tivity along this interstream pathway is dynamically recruited onlywhen both object and spatial information are processed in the brain.Thus, the functional connections together with the anatomical connec-tions form a flexible network for supporting both segregation and inte-gration of the dorsal and ventral visual streams.

Limitation of the current study

In the current study design, it was possible that the spatial taskcould have induced micro saccades. Micro saccades have beenreported to cause similar activation patterns as small visually guidedsaccades (e.g. Tse et al., 2010), and therefore we cannot exclude thepossibility that the activated regions in the spatial task in the currentstudy could have related to the induced micro saccades. However, aswe showed in the Result section, we did not find activation in the eyemovement related areas. Therefore we suspect that the effect ofeye-movement to the results of working memory was minimal inthis study.

We presented visual stimuli in a block design, because it seemed tobe confusing for subjects to challenge randomized types of trials thatwould require set-shifting when the trial types changed, and becauseset-shifting would in turn require preparation. Even though it wouldbe possible to subtract post hoc the component of set-shifting relatedactivation, the task itself would be confusing to the subjects, and theerror rate would likely increase as a direct result of set-shifting, whichis not a focus of this study. The correct trials were 100% for the simple

face and spatial tasks, and we only used the activated areas in thetasks with 100% performance as seed regions to initiate tractography.Therefore, the error did not affect our main results (tractography).However, the activation in the combined task would contain not onlyworking memory but also an error-monitoring component.

The spatial and object tasks differ not only in spatial versus object(face) processing but also in the eccentricity of the stimuli, and thus,in principle it is possible that this eccentricity difference is partially re-sponsible for the differential activation. However, the spatial stimuliused in this studywere located at a visual angle 2.5° away from the cen-ter, which was within the foveal representation (2.6°), and could driveventral streams in the temporal cortex, as well as dorsal streams in theparietal cortex (Hasson et al., 2002; Levy et al., 2001). The nature of thestimuli was also different between tasks: a simple dot versus a face.However, we used the matched stimuli in control tasks, and thereforethe effect of the stimulus difference should be mostly removed bycontrol tasks that we used for each condition.

Acknowledgments

We thank Nichole Eusemann for editorial support. This work wassupported by the National Institutes of Health (R01NS44825) (DSK);the Human Frontiers Science Program; and the Uehara MemorialFoundation (Japan). This work was also partly supported by the EuniceShriver Kennedy National Institute of Child Health and Development(R21HD069001) (ET). This researchwas carried out in part at theAthinoulaA. Martinos Center for Biomedical Imaging at the Massachusetts GeneralHospital, using resources provided by the Center for FunctionalNeuroimaging Technologies, P41RR14075, a P41 Regional Resourcesupported by the Biomedical Technology Program of the NationalCenter for Research Resources (NCRR), National Institutes of Health.

Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.neuroimage.2012.10.002.

References

Alexander, A.L., Hasan, K.M., Lazar, M., Tsuruda, J.S., Parker, D.L., 2001. Analysis of par-tial volume effects in diffusion-tensor MRI. Magn. Reson. Med. 45, 770–780.

Andersen, R.A., Asanuma, C., Cowan, W.M., 1985. Callosal and prefrontal associationalprojecting cell populations in area 7A of the macaque monkey: a study using retro-gradely transported fluorescent dyes. J. Comp. Neurol. 232, 443–455.

Basser, P.J., Mattiello, J., LeBihan, D., 1994. Estimation of the effective self-diffusion tensorfrom the NMR spin echo. J. Magn. Reson. Med. 103, 247–254.

Basser, P.J., Pajevic, S., Pierpaoli, C., Duda, J., Aldroubi, A., 2000. In vivo fibertractography using DT-MRI data. Magn. Reson. Med. 44, 625–632.

Barash, S., Bracewell, R.M., Fogassi, L., Gnadt, J.W., Andersen, R.A., 1991a. Saccade-relatedactivity in the lateral intraparietal area. I. Temporal properties; comparison witharea 7a. J. Neurophysiol. 66, 1095–1108.

Barash, S., Bracewell, R.M., Fogassi, L., Gnadt, J.W., Andersen, R.A., 1991b. Saccade-relatedactivity in the lateral intraparietal area. II. Spatial properties. J. Neurophysiol. 66,1109–1124.

Barbas, H., 1988. Anatomic organization of basoventral and mediodorsal visual recipientprefrontal regions in the rhesus monkey. J. Comp. Neurol. 276, 313–342.

Barbas, H., Mesulam, M.M., 1985. Cortical afferent input to the principalis region of therhesus monkey. Neurosci. 15, 619–637.

Barbas, H., Pandya, D.N., 1989. Architecture and intrinsic connections of the prefrontalcortex in the rhesus monkey. J. Comp. Neurol. 286, 353–375.

Batuev, A.S., Shaefer, V.I., Orlov, A.A., 1985. Comparative characteristics of unit activityin the prefrontal and parietal areas during delayed performance in monkeys.Behav. Brain Res. 16, 57–70.

Behrens, T.E., Johansen-Berg, H., Woolrich, M.W., Smith, S.M., Wheeler-Kingshott, C.A.,Boulby, P.A., Barker, G.J., Sillery, E.L., Sheehan, K., Ciccarelli, O., et al., 2003.Non-invasivemappingof connections between human thalamus and cortex using diffusion imaging.Nat. Neurosci. 6, 750–757.

Belger, A., Puce, A., Krystal, J.H., Gore, J.C., Goldman-Rakic, P., McCarthy, G., 1998. Disso-ciation of mnemonic and perceptual processes during spatial and nonspatial work-ing memory using fMRI. Hum. Brain Mapp. 6, 14–32.

Bruce, C., Desimone, R., Gross, C.G., 1981. Visual properties of neurons in a polysensoryarea in superior temporal sulcus of the macaque. J. Neurophysiol. 46, 369–384.

Catani, M., Thiebaut de Schotten, M., 2009. A diffusion tensor imaging tractographyatlas for virtual in vivo dissections. Cortex 44, 1105–1132.

497E. Takahashi et al. / NeuroImage 65 (2013) 488–498

Cavada, C., Goldman-Rakic, P.S., 1989. Posterior parietal cortex in rhesus monkey: II.Evidence for segregated corticocortical networks linking sensory and limbic areaswith the frontal lobe. J. Comp. Neurol. 287, 422–445.

Chafee, M.V., Goldman-Rakic, P.S., 1998. Matching patterns of activity in primate pre-frontal area 8a and parietal area 7ip neurons during a spatial working memorytask. J. Neurophysiol. 79, 2919–2940.

Cohen, J.D., Perlstein, W.M., Braver, T.S., Nystom, L.E., Noll, D.C., Jonides, J., Smith, E.E.,1997. Temporal dynamics of brain activation during a working memory task. Na-ture 386, 604–608.

Conturo, T.E., Lori, N.F., Cull, T.S., Akbudak, E., Snyder, A.Z., Shimony, J.S., McKinstry,R.C., Burton, H., Raichle, M.E., 1999. Tracking neuronal fiber pathways in the livinghuman brain. Proc. Natl. Acad. Sci. U. S. A. 96, 10422–10427.

Courtney, S.M., Ungerleider, L.G., Keil, K., Haxby, J.V., 1997. Transient and sustained ac-tivity in a distributed neural system for human working memory. Nature 386,608–611.

Courtney, S.M., Ungerleider, L.G., Keil, K., Haxby, J.V., 1996. Object and spatial visualworking memory activate separate neural systems in human cortex. Cereb. Cortex6, 39–49.

Courtney, S.M., Petit, L., Maisog, J.M., Ungerleider, L.G., Haxby, J.V., 1998a. An area spe-cialized for spatial working memory in human frontal cortex. Science 279,1347–1351.

Courtney, S.M., Petit, L., Haxby, J.V., Ungerleider, L.G., 1998b. The role of prefrontal cor-tex in working memory: examining the contents of consciousness. Philos. Trans. R.Soc. Lond. B Biol. Sci. 353, 1819–1828.

Cuzzocreo, J.L., Yassa, M.A., Verduzco, G., Honeycutt, N.A., Scott, D.J., Bassett, S.S., 2009.Effect of handedness on fMRI activation in the medial temporal lobe during an au-ditory verbal memory task. Hum. Brain Mapp. 30, 1271–1278.

Desimone, R., Albright, T.D., Gross, C.G., Bruce, C., 1984. Stimulus-selective properties ofinferior temporal neurons in the macaque. J. Neurosci. 4, 2051–2062.

D'Esposit, M., Aguirre, G.K., Zarahn, E., Ballard, D., Shin, R.K., Lease, J., 1998. FunctionalMRI studies of spatial and non-spatial working memory. Cogn. Brain Res. 7,1–13.

Dougherty, R.F., Ben-Shachar,M., Bammer, R., Brewer, A.A.,Wandell, B.A., 2005. Functionalorganization of human occipito-callosal fiber tracts. Proc. Natl. Acad. Sci. U. S. A. 102,7350–7355.

Felleman, D.J., Van Essen, D.C., 1991. Distributed hierarchical processing in the primatecerebral cortex. Cereb. Cortex 1, 1–47.

Fiez, J.A., Raife, E.A., Balota, D.A., Schwarz, J.P., Raichle, M.E., Petersen, S.E., 1996. A pos-itron emission tomography study of the short-term maintenance of verbal infor-mation. J. Neurosci. 16, 808–822.

Friston, K.J., Stephan, K.E., Lund, T.E., Morcom, A., Kiebel, S., 2005. Mixed-effects andfMRI studies. Neuroimage 24, 244–252.

Friston, K.J., Worsley, K.J., Frackowiak, R.S.J., Mazziotta, J.C., Evans, A.C., 1994. Assessingthe significance of focal activations using their spatial extent. Hum. Brain Mapp. 1,214–220.

Funahashi, S., Bruce, C.J., Goldman-Rakic, P.S., 1989. Mnemonic coding of visual spacein the monkey's dorsolateral prefrontal cortex. J. Neurophysiol. 61, 331–349.

Fuster, J.M., Alexander, G.E., 1971. Neuron activity related to short-term memory.Science 173, 652–654.

Fuster, J.M., Jervey, J.P., 1981. Inferotemporal neurons distinguish and retain behavior-ally relevant features of visual stimuli. Science 212, 952–955.

Gilbert, A.J., Burgess, P.W., 2008. Executive function. Curr. Biol. 18, R110–R114.Gnadt, J.W., Andersen, R.A., 1988. Memory related motor planning activity in posterior

parietal cortex of macaque. Exp. Brain Res. 70, 216–220.Goldman-Rakic, P.S., 1988. Topography of cognition: parallel distributed networks in

primate association cortex. Annu. Rev. Neurosci. 11, 137–156.Goodale, M.A., Milner, A.D., 1992. Separate visual pathways for perception and action.

Trends Neurosci. 15, 20–25.Gross, C.G., Rocha-Miranda, C.E., Bender, D.B., 1972. Visual properties of neurons in

inferotemporal cortex of the Macaque. J. Neurophysiol. 35, 96–111.Guye, M., Parker, G.J., Symms, M., Boulby, P., Wheeler-Kingshott, C.A., Salek-Haddadi,

A., Barker, G.J., Duncan, J.S., 2003. Combined functional MRI and tractography todemonstrate the connectivity of the human primary motor cortex in vivo.Neuroimage 19, 1349–1360.

Hasson, U., Levy, I., Behrmann, M., Hendler, T., Malach, R., 2002. Eccentricity bias as anorganizing principle for human high-order object areas. Neuron 34, 479–490.

Hatta, T., 2007. Handedness and the brain: a review of brain-imaging techniques.Magn. Reson. Med. Sci. 6, 99–112.

Haxby, J.V., Ungerleider, L.G., Horwitz, B., Rapoport, S.I., Grady, C.L., 1995. Hemisphericdifferences in neural systems for face working memory: a PET rCBF study. Hum.Brain Mapp. 3, 68–82.

Johansen-Berg, H., Rushworth, M.F.S., 2009. Using diffusion imaging to study humanconnectional anatomy. Annu. Rev. Neurosci. 32, 75–94.

Jones, D.K., Simmons, A., Williams, S.C., Horsfield, M.A., 1999. Non-invasive assessmentof axonal fiber connectivity in the human brain via diffusion tensor MRI. Magn.Reson. Med. 42, 37–41.

Kaas, J.H., 2004. Early visual areas: V1, V2, V3, DM, DL, andMT. In: Kaas, J.H., Collins, C.E.(Eds.), The Primate Visual System. CRC Press, Boca Raton, pp. 139–159.

Kawamura, K., Naito, J., 1984. Corticocortical projections to the prefrontal cortex in therhesus monkey investigated with horseradish peroxidase techniques. Neurosci.Res. 1, 89–103.

Kim, M., Ducros, M., Carlson, T., Ronen, I., He, S., Ugurbil, K., Kim, D.-S., 2006. Anatom-ical correlates of the functional organization in the human occipitotemporal cortex.Magn. Reson. Imaging 24, 583–590.

Koch, K.W., Fuster, J.M., 1989. Unit activity in monkey parietal cortex related to hapticperception and temporary memory. Exp. Brain Res. 76, 292–306.

Konen, C.S., Kastner, S., 2008. Two hierarchically organized neural systems for objectinformation in human visual cortex. Nat. Neurosci. 11, 224–231.

Kubota, K., Niki, H., 1971. Prefrontal cortical unit activity and delayed alternation per-formance in monkeys. J. Neurophysiol. 34, 337–347.

Lachica, E.A., Beck, P.D., Casagrande, V.A., 1992. Parallel pathways inmacaquemonkey stri-ate cortex: anatomically defined columns in layer III. Proc. Natl. Acad. Sci. U. S. A. 89,3566–3570.

Lazar, M., Alexander, A.L., 2005. Bootstrap white matter tractography (BOOT-TRAC).Neuroimage 24, 524–532.

Lazar, M., Weinstein, D.M., Tsuruda, J.S., Hasan, K.M., Arfanakis, K., Meyerand, M.E.,Badie, B., Rowley, H.A., Haughton, V., Field, A., Alexander, A.L., 2003. White mattertractography using diffusion tensor deflection. Hum. Brain Mapp. 18, 306–321.

Lehericy, S., Ducros, M., Krainik, A., Francois, C., Van de Moortele, P.F., Ugurbil, K., Kim,D.-S., 2004. 3-D diffusion tensor axonal tracking shows distinct SMA and pre-SMAprojections to the human striatum. Cereb. Cortex 14, 1303–1309.

Levy, R., Goldman-Rakic, P.S., 2000. Segregation of working memory functions withinthe dorsolateral prefrontal cortex. Exp. Brain Res. 133, 23–32.

Levy, I., Hasson, U., Avidan, G., Hendler, T., Malach, R., 2001. Center-periphery organiza-tion of human object areas. Nat. Neurosci. 4, 533–539.

Livingstone, M., Hubel, D., 1988. Segregation of form, color, movement, and depth:anatomy, physiology, and perception. Science 240, 740–749.

Lyon, D.C., 2007. The evolution of visual cortex and visual systems. In: Kaas, J.H., Krubitzer,L.A. (Eds.), Evolution of Nervous Systems, vol. 3. Elsevier, London, pp. 267–306.

McCarthy, G., Puce, A., Constable, R.T., Krystal, J.H., Gore, J.C., Goldman-Rakic, P., 1996.Activation of human prefrontal cortex during spatial and nonspatial workingmemory tasks measured by functional MRI. Cereb. Cortex 6, 600–611.

McIntosh, A.R., Grady, C.L., Ungerleider, L.G., Haxby, J.V., Rapoport, S.I., Horwitz, B.,1994. Network analysis of cortical visual pathways mapped with PET. J. Neurosci.14, 655–666.

Mellet, E., Tzourio, N., Crivello, Joliot, Dnis,M.,Mazoyer, B., 1996. Functional anatomy of spa-tial mental imagery generated from verbal instructions. J. Neurosci. 16, 6504–6512.

Mesulam, M.M., Van Hoesen, G.W., Pandya, D.N., Geschwind, N., 1977. Limbic and sen-sory connections of the inferior parietal lobule (area PG) in the rhesus monkey: astudy with a new method for horseradish peroxidase histochemistry. Brain Res.136, 393–414.

Miller, E.K., Li, L., Desimone, R., 1991. A neural mechanism for working and recognitionmemory in inferior temporal cortex. Science 254, 1377–1379.

Miyashita, Y., Chang, H.S., 1988. Neuronal correlate of pictorial short-term memory inthe primate temporal cortex. Nature 331, 68–70.

Mohr, H.M., Goebel, R., Linden, D.E., 2006. Content- and task-specific dissociations offrontal activity during maintenance and manipulation in visual working memory.J. Neurosci. 26, 4465–4471.

Mori, S., 2007. Introduction to Diffusion Tensor Imaging. Elsevier Science.Mori, S., Crain, B.J., Chacko, V.P., van Zijl, P.C., 1999. Three-dimensional tracking of axonal

projections in the brain by magnetic resonance imaging. Ann. Neurol. 45, 265–269.Nassi, J.J., Callaway, E.M., 2009. Parallel processing strategies of the primate visual sys-

tem. Nat. Rev. Neurosci. 10, 360–372.Nystrom, L.E., Bracer, T.S., Sabb, F.W., Delgado, M.R., Noll, D.C., Cohen, J.D., 2000. Work-

ing memory for letters, shapes, and locations: fMRI evidence against stimulus-based regional organization in human prefrontal cortex. Neuroimage 11, 424–446.

O'Scalaidhe, S.P., Wilson, F.A., Goldman-Rakic, P.S., 1997. Areal segregation of face-processing neurons in prefrontal cortex. Science 278, 1135–1138.

Owen, A.M., Evans, A.C., Petrides, M., 1996. Evidence for a two-stage model of spatialworking memory processing within the lateral frontal cortex: a positron emissiontomography study. Cereb. Cortex 6, 31–38.

Owen, A.M., 1997. The functional organization of working memory processes withinhuman lateral frontal cortex: the contribution of functional neuroimaging. Eur. J.Neurosci. 9, 1329–1339.

Petit, L., Orssaud, C., Petit, L., Orssaud, C., tzourio, N., Crivello, F., Berthoz, A., Mazoyer, B.,1996. Functional anatomy of a prelearned sequence of horizontal saccades inhumans. J. Neurosci. 16, 3714–3726.

Petrides, M., Pandya, D.N., 1984. Projections to the frontal cortex from the posterior pa-rietal region in the rhesus monkey. J. Comp. Neurol. 228, 105–116.

Petrides, M., Tomaiuolo, F., Yeterian, E.H., Pandya, D.N., 2012. The prefrontal cortex:comparative architectonic organization in the human and the macaque monkeybrains. Cortex 48, 46–57.

Postle, B.R., Stern, C.E., Rosen, B.R., Corkin, S., 2000. An fMRI investigation of corticalcontributions to spatial and nonspatial visual working memory. Neuroimage 11,409–423.

Powell, H.W., Guye, M., Parker, G.J., Symms, M.R., Boulby, P., Koepp, M.J., Duncan, J.S.,G.J., Barker, 2004. Noninvasive in vivo demonstration of the connections of thehuman parahippocampal gyrus. Neuroimage 22, 740–747.

Power, J.D., Barnes, K.A., Snyder, A.Z., Schlaggar, B.L., Petersen, S.E., 2012. Spurious butsystematic correlations in functional connectivity MRI networks arise from subjectmotion. Neuroimage 59, 2142–2154.

Propper, R.E., O'Donnell, L.J., Whalen, S., Tie, Y., Norton, I.H., Suarez, R.O., Zollei, L.,Radmanesh, A., Golby, A.J., 2010. A combined fMRI and DTI examination of func-tional language lateralization and arcuate fasciculus structure: effects of degreeversus direction of hand preference. Brain Cogn. 73, 85–92.

Rao, S.C., Rainer, G., Miller, E.K., 1997. Integration of what and where in the primateprefrontal cortex. Science 276, 821–824.

Sala, J.B., Courtney, S.M., 2007. Binding of what and where during working memorymaintenance. Cortex 43, 5–21.

Sato, T., Kawamura, T., Iwai, E., 1980. Responsiveness of inferotemporal single units to visualpattern stimuli in monkeys performing discrimination. Exp. Brain Res. 38, 313–319.

Schmahmann, J.D., Pandya, D.N., 2006. Fiber Pathways of the Brain. Oxford University Press.

498 E. Takahashi et al. / NeuroImage 65 (2013) 488–498

Schmahmann, J.D., Pandya, D.N., Wang, R., Dai, G., D'Arceuil, H.E., de Crespigny, A.J.,Wedeen, V.J., 2007. Association fibre pathways of the brain: parallel observationsfrom diffusion spectrum imaging and autoradiography. Brain 130, 630–653.

Schneider, G.E., 1969. Two visual systems. Science 163, 895–902.Sincich, L.C., Horton, J.C., 2002. Divided by cytochrome oxidase: a map of the projec-

tions from V1 to V2 in macaques. Science 295, 1734–1737.Smith, E.E., Jonides, J., Koeppe, R.A., Awh, E., Schumacher, E.H., Minoshima, S., 1995. Spatial

versus object working memory: PET investigations. J. Cogn. Neurosci. 7, 337–356.Stieltjes, B., Kaufmann, W.E., van Zijl, P.C., Fredericksen, K., Pearlson, G.D., Solaiyappan,

M., Mori, S., 2001. Diffusion tensor imaging and axonal tracking in the humanbrainstem. Neuroimage 14, 723–735.

Takahashi, E., Ohki, K., Kim, D.S., 2007. Diffusion tensor studies dissociated two fronto-temporal pathways in the human memory system. Neuroimage 34, 827–838.

Takahashi, E., Ohki, K., Kim, D.S., 2008. Dissociated pathways for successful memory re-trieval from the human parietal cortex: anatomical and functional connectivityanalyses. Cereb. Cortex 18, 1771–1778.

Talairach, J., Tournoux, P., Rayport, M., 1988. Co-planar Stereotaxic Atlas of the Human Brain:3-D Proportional System: An Approach to Cerebral Imaging. Georg Thieme Verlag.

Thiebaut de Schotten, M., Urbanski, M., Duffau, H., Volle, E., Levy, R., Dubois, B.,Bartolomo, P., 2005. Direct evidence for a parietal–frontal pathway subserving spa-tial awareness in humans. Science 9, 2226–2228.

Thottakara, P., Lazar, M., Johnson, S.C., Alexander, A.L., 2006. Application of Brodmann'sarea templates for ROI selection in white matter tractography studies. Neuroimage29, 868–878.

Toosy, A.T., Ciccarelli, O., Parker, G.J., Wheeler-Kingshott, C.A., Miller, D.H., Thompson, A.J.,2004. Characterizing function–structure relationships in the human visual systemwith functional MRI and diffusion tensor imaging. Neuroimage 21, 1452–1463.

Tse, P.U., Baumgartner, F.J., Greenlee, M.W., 2010. Event-related functional MRI of cor-tical activity evoked by microsaccades, small visually-guided saccades, andeyeblinks in human visual cortex. Neuroimage 49, 805–816.

Ungerleider, L.G., Mishkin, M., 1982. Object vision and spatial vision: two cortical path-ways. In: Ingle, D.J., Goodale, M.A., Mansfield, R.J.W. (Eds.), Analysis of Visual Be-havior. MIT Press, pp. 549–586.

Ungerleider, L.G., Haxby, J.V., 1994. ‘What’ and ‘where’ in the human brain. Curr. Opin.Neurobiol. 4, 157–165.

Urbanski, M., et al., 2008. Brain networks of spatial awareness: evidence from diffusiontensor imaging tractography. J. Neurol. Neurosurg. Psychiatry 79, 598–601.

Volle, E., Kinkingnehun, S., Pochon, J.-B., Mondon, K., Thiebaut de Schotten, M., Seassau,M., Duffau, H., Samson, Y., Dubois, B., Levy, R., 2008. The functional architecture ofthe left posterior and lateral prefrontal cortex in humans. Cereb. Cortex 18,2460–2469.

Wilson, F.A., O'Scalaidhe, S.P., Goldman-Rakic, P.S., 1993. Dissociation of object andspatial processing domains in primate prefrontal cortex. Science 260, 1955–1958.

Xu, D., Mori, S., Solaiyappan, M., van Zijl, P.C., Davatzikos, C., 2002. A framework forcallosal fiber distribution analysis. Neuroimage 17, 1131–1143.

Yabuta, N.H., Callaway, E.M., 1998. Functional streams and local connections of layer 4Cneurons in primary visual cortex of the macaque monkey. J. Neurosci. 18,9489–9499.