NeuroImage 52 (2010) 1230–1237

Contents lists available at ScienceDirect

NeuroImage

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

Structural connectivity of Broca's area and medial frontal cortex

Anastasia Ford a,b,⁎, Keith M. McGregor a,b, Kimberly Case a,b, Bruce Crosson a,c, Keith D. White a,b,c

a Brain Rehabilitation Research Center, Malcom Randall VA Medical Center, Gainesville, FL 32608, USAb Department of Psychology, University of Florida, Gainesville, FL 32611, USAc Department of Clinical and Health Psychology, University of Florida, Gainesville, FL 32610, USA

⁎ Corresponding author. Department of PsychologyBox 112250, Gainesville, FL 32611-2250, USA. Fax: +1

E-mail address: sokolova@ufl.edu (A. Ford).

1053-8119/$ – see front matter © 2010 Elsevier Inc. Adoi:10.1016/j.neuroimage.2010.05.018

a b s t r a c t

a r t i c l e i n f o

Article history:Received 25 August 2009Revised 2 May 2010Accepted 7 May 2010Available online 20 May 2010

Despite over 140 years of research on Broca's area, the connections of this region to medial frontal cortexremain unclear. The current study investigates this structural connectivity using diffusion-weighted MRItractography in living humans. Our results show connections between Broca's area and Brodmann's areas(BA) 9, 8, and 6 (both supplementary motor area (SMA) in caudal BA 6, and Pre-SMA in rostral BA 6).Trajectories follow an anterior-to-posterior gradient, wherein the most anterior portions of Broca's areaconnect to BA 9 and 8 while posterior Broca's area connects to Pre-SMA and SMA. This anterior–posteriorconnectivity gradient is also present when connectivity-based parcellation of Broca's area is performed.Previous studies of language organization suggest involvement of anterior Broca's area in semantics andposterior Broca's area in syntax/phonology. Given corresponding patterns of functional and structuralorganization of Broca's area, it seems well warranted to investigate carefully how anterior vs. posteriormedial frontal cortex differentially affect semantics, syntax and phonology.

, University of Florida, P. O.352 392 7985.

ll rights reserved.

© 2010 Elsevier Inc. All rights reserved.

Introduction

Despite over 140 years of research on Broca's area, the connectionsof this region to medial frontal cortex in human brain remain unclear.Classical views of language organization, based primarily on humanbrain lesions, have long implicated Broca's and Wernicke's areas asimportant cortical loci of language processing (Broca, 1861;Wernicke,1874; Lichtheim, 1885). More recent studies have identified furthercortical substrates, as well as subcortical structures, involved inlanguage (Binder et al., 1997; Crosson, 1999; Crosson et al., 2007)building upon the development of functional neuroimaging techni-ques such as positron emission tomography and functional magneticresonance imaging (fMRI). Many of these studies have identifiedregions that co-activate during language paradigms, and thus arethought to be members of the same functional network (Crossonet al., 2001; Horwitz et al., 1998). While functional connectivity asevidenced by co-activation implies the existence of a neural network,it does not define structural connectivity per se. Known structuralconnectivity of a given neural substrate can provide insight into thephysiological mechanisms for network organization, as well as intopotential functional characteristics.

Much of what is currently known about structural connectivity ofhuman language areas comes indirectly from non-human primateliterature. Comparative cytoarchitectonic studies have identified

monkey homologues of human Broca's region, namely Brodmann'sareas (BA) 44 and 45 (Petrides and Pandya, 1994; Petrides, 2005). Inparticular, using architectonic comparison and electrophysiologicalrecordings and microstimulation, Petrides and colleagues found anarea anterior to monkey area 44 involved with control of orofacialmusculature (Petrides, 2005). The authors postulated that this areafurther developed to control communicative acts and eventuallyhuman speech in the progression of primate phylogeny. In addition,connectivity patterns of monkey BA 44/45 are distinct, implyingdifferences in their functional domain. While area 44 is connectedwith the ventral premotor cortex, which has orofacial and hand/armrepresentations, BA 45 in monkey is connected to the superiortemporal sulcus (STS). Further work by Petrides and Pandyademonstrated projections from this monkey analogue of Broca'sregion to the superior temporal gyrus (STG), the upper bank of STS(monkey analogues of humanWernicke's area), and also to themedialfrontal cortex (Petrides and Pandya, 2002).

Projections from human BA 44/45 to the STG have been implied bya number of fMRI studies suggesting functional connectivity (Kleinet al., 1995; Petrides, 1995; Binder et al., 1997) as well as by diffusion-weighted MRI (DW-MRI) suggesting structural connectivity (SeeFriedrici, 2009 for a comprehensive review). Recent studies employ-ing applications of DW-MRI have provided many valuable insightsinto Broca's area connectivity. Catani and colleagues (2005) demon-strated direct as well as indirect connections between Broca's andWernicke's areas serving phonological and semantic functions,respectively. Glasser and Rilling divided the arcuate fasciculus intotwo segments: one originating in BA 44/45 and terminating in the

Table 1Age, gender, and education history of the participants.

Participant ID Age Gender Years of education

S1 22 F 16S2 25 M 16S3 32 F 16S4 64 F 16S5 65 F 14S6 70 F 16S7 74 F 13S8 81 M 20S9 82 F 18

1231A. Ford et al. / NeuroImage 52 (2010) 1230–1237

posterior STG subserving phonology, and the other originating in BA44/45 and BA 6/9 and terminating in the middle temporal gyrus(MTG) subserving lexical-semantic processing (Glasser and Rilling,2008).

Functional connectivity of Broca's area with medial frontal cortexhas been inferred from a number of fMRI studies (Crosson et al., 2001;Binkofski et al., 2000), as well as being strongly implied by the ‘mirror-neuron’ theory of language (Rizzolatti and Craighero, 2004). Struc-tural connectivity with the medial frontal cortex has been shownpreviously, although not studied extensively (Anwander et al., 2007,p. 820). Better understanding of the connectivity between theseregions will provide valuable insights into language organization as itmay relate to the selection and initiation of motor acts, and to ‘mirror-neuron’ sensorimotor networks which are present in monkey BA 44/45 (Rizzolatti and Arbib, 1998). Clinically, the existence of theseconnections and knowledge about their trajectories could help tobetter predict language impairments and recovery after lesionsinvolving Broca's area, or white matter in its vicinity or along thesetrajectories, or medial cortex to which Broca's area is connected.Understanding these connections also eventually can be useful inspecifying the roles of medial frontal cortex in language functions.Medial frontal cortex has previously been implicated in actionselection/outcome monitoring, behavioral adjustments, and learning(Ridderinkhof et al., 2004; Rushworth et al., 2004, 2007), all of whichare important in complex cognitive tasks such as language. Monitor-ing selection and retrieval of grammatically correct and contextuallyappropriate responses, as well as adjusting behavioral output during alanguage task, would suggest a logical role for medial frontal cortexextending beyond classical language functions.

Structural connectivity between themonkey homologue of Broca'sarea and the medial frontal cortex has been previously demonstrated(Petrides and Pandya, 2002). To infer similar connectivity in humans,the present study identified analogous cortical regions in the medialfrontal cortex as projection sites of Broca's area, namely BA 6(caudally, as supplementary motor area (SMA); rostrally, as Pre-SMA), BA 8, and BA 9. The supracallosal BA 32 was included in thepresent analysis by extending the inferior boundary of the medialfrontal masks to the cingulate sulcus. Because the paracingulate sulcusis absent in a large minority of humans, the borders between BA 8 or 6and BA 32 are difficult to establish in those subjects (Crosson, 1999).Supracallosal BA 32 is thought to be equivalent to the cingulate motorareas in macaques. Because the cingulate motor areas in macaqueshave similar connectivity to their superior and adjacent counterpartsin medial BA 6 (i.e., SMA, Pre-SMA), it seems reasonable to include BA32 within the masks, although it is not specifically referenced in ournaming conventions for these masks.

The present study employs DW-MRI to infer structural connectiv-ity of human BA 44/45 specifically to medial frontal cortex. DW-MRIallows in-vivo visualization of white matter in the brain as inferredfrom the directional diffusion of water (Basser et al., 1994). The basicrationale is that, over a few tens of milliseconds, water molecules atnormal physiological temperature can travel equal distances in anydirection (i.e., isotropically) when they are in the middle of a largecompartment like a ventricle, but can only travel more readily inselect directions (i.e., anisotropically) if bounded by featuresimpermeable to water like lipid cell membranes and myelin sheaths.Within a particular DW-MRI voxel, the preponderance of anisotropicover isotropic diffusion characterizes the voxel's fractional anisotropy(FA). The predominant direction of the anisotropic diffusion char-acterizes the voxel's principal eigenvector (PE). The quantities of FAand PE are determined by fitting a smooth tensor to DW-MRI dataobtained by indexing diffusion along many directions. When PEshaving nearly the same direction align or seemingly connect acrosscontiguous voxels, they form a streamline suggestive of a tract.

The present study inferred tracts originating in BA 44/45 usingprobabilistic tractography, a fiber-tracking algorithm that provides

better visualization of branching and crossing fibers than does a moretraditional streamlining algorithm, and it also generates empiricallikelihood estimates associatedwith the inferred tracts (Behrens et al.,2007). We also parcellated Broca's area using voxelwise likelihoods ofconnections originating in BA 44/45 and connecting to each of theaforementioned regions of medial cortex (Behrens et al., 2003b;Anwander et al., 2007). Probabilistic tractography enabled us tovisualize trajectories of projection between Broca's area and regions inthe medial frontal cortex, while connectivity-based segmentationallowed us to segment Broca's area based on the likelihood ofconnection to each of the four targets.

Methods

Participants

Nine right-handed, native English speakers, with no reportedneurological disorders were recruited. Table 1 shows gender, age, andeducation level for each participant. Three out of nine of our participantswere in the age range of 22–32 (mean=26.3, st.dev.=5.13) andwill bereferred to as younger participants. Six out of nine of our participantswere in the age range of 64–82 (mean=72.67, st.dev.=7.74) and willbe referred to as older participants. Written informed consent wasobtained from all participants in compliance with Institutional ReviewBoard guidelines of the University of Florida and the North Florida/South Georgia Malcom Randall Veteran's Affairs Medical Center.

Image acquisition

All scans were obtained for each participant on a Philips Achieva3 T scanner (Amsterdam, Netherlands) using an 8-channel SENSEhead coil. After a three-plane localizer scout used to center the fieldsof view to the participant's brain, and a field sensitivity reference scanfor SENSE image reconstruction, a structural MP-RAGE T1-weightedscan was acquired with 130×1.0 mm sagittal slices, FOV=240 mm(AP)×180 mm (FH), matrix=256×192, TR=9.90 ms, TE=4.60 ms,FA=8°, voxel size=1.0 mm×0.94 mm×0.94 mm,, and time ofacquisition=6 min 11 s. Diffusion-weighted images were acquiredusing single shot spin-echo echo planar imaging (EPI) with60×2.0 mm axial slices (no gap), FOV=224 mm (AP)×224 mm(RL), matrix=112×112, TR=9509 ms, TE=55 ms, FA=90°, voxelsize=2.0×2.0×2.0 mm, and time of acquisition=5 min 42 s.. Thediffusion weighing gradients were isotropically distributed in spaceusing a 33-direction acquisition scheme with b=1000 s/mm2. Asingle volume with no diffusion weighting (b=0) was also acquiredwith these parameters.

Data analysis

Data processing was performed using the FSL software packageFMRIB software Library (www.fmrib.ox.ac.uk/fsl). First, the eddycurrent distortions were removed (FDT, version 2.0), then eachdiffusion direction volume was registered to the b=0 volume using

1232 A. Ford et al. / NeuroImage 52 (2010) 1230–1237

FLIRT (Jenkinson and Smith, 2001). Next, a binary mask of brain-onlyvoxels was generated from the b=0 volume (Smith, 2002). Thestructural scans of each participant were skull stripped (Smith, 2002)and then registered to the participant's diffusion space (Jenkinson andSmith, 2001).

The tensor fit was performed in brain-only diffusion voxels(Behrens et al., 2003a). The fractional anisotropy (FA) and theprinciple eigenvector (PE) maps were examined to ensure proper fitand absence of distortions. Next, distributions of diffusion parameters,such as in-voxel fiber orientation, were estimated by running MarkovChain Monte Carlo (MCMC) sampling in each brain-only diffusionvoxel (Behrens et al., 2007). Probabilistic tractography used thesedistributions to infer tracts connecting the seed voxels of interest inBroca's area to the targets in medial frontal cortex.

TractographyInferred fibers connecting voxels defined by a Broca's area seed

mask to each of themedial frontal cortex target maskswere visualized(Behrens et al., 2007). Details of the masks are given below. Thistracking algorithm first estimates voxelwise probability densitydistributions for each parameter of the two-compartment simplepartial volume model of diffusion, modeling 2 crossing fibers pervoxel (Behrens et al., 2007) using MCMC. The two-compartmentmodel assumes that all of the anisotropy in a given voxel comes fromthe presence of an inferred fiber, whereas all isotropic componentswould come from water surrounding the fiber. MCMC empiricallyestimates probability densities associated with inferred fiber orienta-tions by drawing 1000 samples from each diffusion voxel. Highlyanisotropic voxels support narrow orientation distributions and moreisotropic voxels support broader distributions, thus no FA threshold isrequired. The tracking algorithm then calculates 5000 probabilisticstreamlines, starting from each diffusion voxel within the seed mask,using step length of 0.5 mm (four steps per voxel), and utilizing theprobability density of local fiber orientations in each voxel encoun-tered along each streamline. Higher local probabilities in adjacentvoxels for the particular fiber orientation that connects them (within acertain tolerance for curvature) will yield a large expected value forthe number of associated streamlines passing through them, incontrast to a low expected value if the most probable orientations inthese voxels are discrepant. We used a curvature threshold of 0.2,which is the cosine of the allowable angle between two steps, as thetolerance parameter of the tracking algorithm.

The number of streamlines traversing each seed-to-target con-nection provides an empirical index concatenating the local orienta-tion distributions of all the voxels traversed by those streamlines.Accumulating these indices for each target across voxels contained inthe seed mask yields a connectivity distribution. Connectivitydistributions for each of the medial frontal areas of interest representhowmany probabilistic streamlines originated in voxels contained bythe Broca's area seed mask, and reached voxels within each of themedial frontal target masks. Seed voxels having 10 or fewer out of5000 possible streamlines connected to the target mask were found tobe spatially disbursed away from the high connectivity voxel clustersof each connectivity distribution and were, therefore, attributed tonoise and excluded from further analysis (see Ciccarelli et al., 2006and Heiervang et al., 2006 for similar approach).

Connectivity-based parcellation

After such exclusions, four sub-regions of interest became evidentwithin the Broca's area (BA 44/45) seed mask. We performed aclassification target analysis using the same four target masks tocompute the likelihoods of connection between each voxel in theBroca's area mask and each of the four medial frontal targets (Behrenset al., 2003b). As a result, each voxel in the Broca's mask had fourassociated connectivity indices, indicating the number of probabilistic

streamlines connecting that voxel to each of the four targets. Weparcellated the Broca's area mask to assign each seed voxel in themask to one of the medial cortical targets based on the highestconnectivity index.

Seeding again from these parcellated sub-regions we calculatednew probabilistic tractography streamlines between each voxel in theBroca's area sub-regions and each medial frontal target separately.Resulting connectivity distributions were truncated to exclude voxelshaving fewer than 1% of the maximum number of streamlinesoriginating within the parcellated seed mask, which number variedacross target masks and across participants. Post-parcellation tracto-graphy is reported in the Results section.

Seed and target masks for tractographyThe masks of Broca's area were drawn on the structural scan of

each subject in the native acquisition space. The lateral-most sagittalslice of the frontal cortex of the skull-stripped structural scan wasused as the lateral border of each Broca's area mask. The medialborder was 10 mm from the lateral border. The Broca's mask's dorsalborder was the inferior frontal sulcus, while the ventral border wasthe Sylvian fissure. The Broca's mask's anterior border was defined bya coronal plane through the anterior margin of the anterior horizontalramus of the Sylvian fissure, and its posterior border was inferiorprecentral sulcus. To define the target regions in medial frontal cortex,the skull-stripped structural scans of each participant were normal-ized to the Talairach atlas (Talairach and Tournoux, 1988) using AFNI(Cox, 1996). The coronal plane perpendicular to the anteriorcommissure–posterior commissure (AC-PC) line, and tangent to theAC caudally in the most medial slice of the left hemisphere, was usedto divide Pre-SMA from SMA within BA 6 (Picard and Strick, 2001).The caudal border of SMA was located 20 mm posterior to thisdivision. The rostral border of Pre-SMA was defined to be 20 mmanterior to this division. Medial BA 8 was defined as extending 20 mmanteriorly from the rostral border of Pre-SMA. The caudal border of BA9 began just anterior to the rostral border of BA 8, while the rostralborder of BA 9 was drawn 20 mm anterior to its caudal border. Allmedial frontal masks were drawn to be non-overlapping andextended 7 slices (∼7 mm) laterally from the medial extent of theleft hemisphere in the Talairach space. The cingulate sulcus served asthe inferior border for each of the medial frontal masks, with the crestof the brain as the superior border. The inferior border of the BA 9mask was defined by a line connecting the genu of the corpuscallosum with the genu of the cingulate sulcus, projecting anteriorlywhere the cingulate sulcus was no longer inferior to BA 9. Each of themedial frontal target masks was back-transformed from Talairachspace into the participant's native acquisition space, and each maskwas then registered to the participant's diffusion space using thelinear transform generated by the registration of the participant'sstructural scan to their diffusion scan (using FSL's FLIRT).

Exclusion masks for tractographyTo prevent potential inclusion of the U-shaped short association

fibers connecting inferior frontal gyrus to adjacent gyri, exclusionmasks were drawn contiguous to Broca's area seed masks to excludeposterior precentral gyrus and superior-anterior middle frontal gyrus.Exclusion masks were also drawn in the corpus callosum and internalcapsule to exclude crossing or descending motor projections.

Analysis of surrogate dataTo evaluate whether these methods would artifactually influence

the inferred anatomy of our participants we created a surrogate dataset (Theiler et al., 1992). First, we generated a 4-dimensional array (X,Y,Z,Gradient Direction) having the geometry of brain-only voxelsfrom participant #1 (Smith, 2002). Next, we inserted randomGaussian noise in each array cell (surrogate voxel). We scaled theresulting noise-only surrogate data to match the range of signal

1233A. Ford et al. / NeuroImage 52 (2010) 1230–1237

intensities of the participant's brain image. Next, we estimateddiffusion parameters of these surrogate data using MCMC (Behrenset al., 2007). We re-ran our tractography analysis using the same seed,termination, and exclusion masks to trace fibers between Broca's areaand the medial frontal cortex, as though the surrogate data were abrain image rather than random noise. Furthermore, we used thecentral voxel of the Broca's area mask to trace projections originatingfrom this location without applying any additional restrictions. Zeroprojections resulted from applying our original method to the randomnoise surrogate data. Inferred “tracts” from the central voxel tracingdispersed only a short distance (b10 voxels) from the startinglocation.

Results

Visualization of connectivity-based parcellation of Broca's area,shown in Fig. 1, shows generally anterior-to-posterior orderedconnectivity. Anterior portions of Broca's area showed the highestlikelihood of connection with anterior medial cortex (BA 9 and/or BA8, light green and dark green respectively), whereas the mostposterior portions of Broca's area typically have posterior medialcortex as their most likely targets for connection (Pre-SMA and/orSMA, blue and orange respectively).

Overall, BA 8 and 9 dominate connectivity in more than 50% ofBroca's area in seven of nine participants (i.e., except for participants 5and 7). Relatively few voxels in Broca's area in any of the participantsshowedhigh likelihoods of connectionwith SMA. For ourfive youngestparticipants (1, 2, 3, 4, 5), aged 22–65 years, no voxel in Broca's areaexhibited an SMA connectivity index higher than those of the

Fig. 1. Connectivity-based parcellation of Broca's area. Light gre

remaining targets in medial frontal cortex. For these five participants,Broca's area parcellated into only three connectivity regions: BA 9, 8,and Pre-SMA. The absence of SMA in Broca's area parcellationwith thepresent analysis does not suggest that Broca's area–SMA connectivitydoes not exist, but rather that SMA has a lower connectivity indexcompared to the other three areas in the medial frontal cortex. Basedon the previous macaque tracer studies of connectivity betweenventral lateral premotor cortex and SMA, if we had included ventrallateral BA 6 and BA 47 in our analysis to determine connectivity of thelarger Broca's region, rather than the smaller BA 44/45 Broca's area,then stronger connectivity with the SMA would be expected (Ghoshand Gattera, 1995). Parcellation-based connectivity between Broca'sarea and SMA was more evident for the oldest four out of nine of ourparticipants (6, 7, 8, 9), aged 70–82 years.

Given that the subjects with and without SMA represented in theBroca's area parcellation were so clearly divided on the basis of age, aMann-Whitney U test was applied to determine if there was adifference in the rank order of ages between those groups. The resultsindicate that participants without SMA representation are signifi-cantly younger than those with SMA representation, U=0, p=0.016(two-tailed). This finding is consistent with an anterior-to-posteriorgradient of white matter decline (e.g., Courchesne et al., 2000)accelerated past age 65 (Westlye et al., in press), which could inprinciple more strongly affect BA 9, 8, and pre-SMA, thus allowing lowSMA connectivity to achieve greater relative prominence. It alsoshould be noted that participant 7 in Fig. 1 showed no parcellation forPre-SMA.

Using probabilistic tractography, we traced inferred fiber bundlesconnecting Broca's area to each of the fourmedial frontal targets. Fig. 2

en: BA 9; dark green: BA 8; blue: Pre-SMA; orange: SMA.

Fig. 2. White matter tracts inferred by probabilistic tractography connecting Broca's area (BA 44/45) and medial frontal cortex. Yellow: BA 9; dark green: BA 8; blue: Pre-SMA;orange: SMA. Images obtained from participants indicated by the same number in Table 1.

1234 A. Ford et al. / NeuroImage 52 (2010) 1230–1237

shows an array of 3D representations of these fibers for each of ournine participants. Fiber orientation and trajectory follow a clearanterior–posterior trend, where more anterior portions of the Broca'sarea project to BA 9 and 8 (yellow and green respectively) and moreposterior portions of Broca's area project to Pre-SMA and SMA (blueand orange respectively). The five younger participants who did notshow any high-likelihood connections between Broca's area and SMAwith connectivity-based parcellation (shown in Fig. 1) accordinglylack corresponding inferred fibers with probabilistic tractography,shown in Fig. 2. The four older participants in whom connectivity-based parcellation was revealed for Broca's area and SMA (Fig. 1) alsorevealed inferred fibers with probabilistic tractography (Fig. 2).

Discussion

Structural connectivity of Broca's area and potential functionalimplications thereof have been the topic of many previous studies. Ithas been shown that Broca's area interconnects with the inferiorparietal lobule, and the superior and middle temporal gyri (Catani etal., 2002; Glasser and Rilling, 2008; Frey et al., 2008), consistent withclassical models of language. However, other cortical regions have yetto be shown as projection sites of human Broca's area (BA 44/45). Thepresent paper examines Broca's area connectivity with medial frontalcortex using diffusion-weighted MRI, probabilistic tractography, andconnectivity-based parcellation. Connectivity between these regionsexists in non-human primates;macaques BA 44 and 45were shown toconnect to BA 6, 8, and 9 inmedial frontal cortex (Petrides and Pandya,2002). Cross-species generality between BA 44 and 45 connectivity inmacaque and Broca's area connectivity in humans was recently

demonstrated by separate fiber pathways connecting BA 45 to STGand BA 44 to inferior parietal lobule in macaque which also wereshown to be present in human (Frey et al., 2008). Based on thesefindings, we hypothesized that medial frontal connectivity of BA 44and BA 45 found in macaque would generalize to Broca's area medialfrontal connectivity in humans. To our knowledge, we are the first toshow structural connections between Broca's area and BA 6, 8, and 9 inhuman medial frontal cortex. With the exception of connectionsbetween SMA and Broca's area, the present findings were consistentacross our subjects regardless of their ages.

Although the present study used a moderate number of diffusion-gradient directions, the signal-to-noise ratio of the brain images wassufficient to represent underlying anatomy. Null results from analysisof the random noise surrogate data clearly show that neither the brainimages’ anterior-to-posterior gradient in connectivity-based parcella-tion, nor trajectories of the inferred tracts, arise artifactually from ourmethods. If that were the case, then surrogate data could producesimilar results as the brain images, and this did not occur.

When examining the overall strength of connectivity betweenBroca's area and the four regions of interest in the medial frontalcortex we note that our findings indicate that for five of ourparticipants parcellation-based connectivity between Broca's areaand SMA was very sparse. For these five participants parcellationmaps of Broca's area contain only three sub-regions showing thehighest likelihood of projection to Pre-SMA, BA 8, and BA 9 (Fig. 1,participants 1, 2, 3, 4, and 5). For the other four participants, the sub-region of Broca's area showing highest connectivity to SMA is muchsmaller compared to sub-regions with highest connectivity to Pre-SMA, BA 8, and BA 9. Previous macaque tracer studies show extensive

1235A. Ford et al. / NeuroImage 52 (2010) 1230–1237

connectivity between SMA and the ventral lateral BA 6 of primatemotor cortex (Ghosh and Gattera, 1995; Dancause et al., 2007). Basedon these results from macaque, we believe that expansion of ourregion of interest to include ventral lateral BA 6 and BA 47, so as toconstitute Broca's region as described by Hagoort and colleagues(2006), would result in stronger connectivity with SMA. Changes inthe SMA connectivity could reflect age of the participants. Participantswho did show SMA connectivity based on parcellation of Broca's areawere older (70–82 years of age) than those who did not show SMAconnectivity (22–65). Age-related deterioration in white matter asmeasured by fractional anisotropy (e.g., Salat et al., 2006; Westlyeet al., in press) or volumetry (e.g., Courchesne et al., 2000; Westlyeet al., in press) has been previously reported to affect the prefrontal ormore anterior cortex to a larger extent than other, more posteriorcortical areas. Thus, reduced integrity of white matter connectingBroca's area more strongly affecting anterior portions of the medialfrontal cortex as compared with more posterior SMA, couldpotentially explain the relatively increased prominence of SMAconnectivity in Broca's area parcellations in the four oldest partici-pants. However, formal analysis of this trend is made difficult by oursmall sample size. Further studies investigating potential age effectson connectivity of Broca's area and the prefrontal cortex should becarried out to fully understand changes in these white matter fibertracts with age.

Previous work on connectivity-based parcellation of Broca's area byAnwanderand colleagues seems to followknownanatomical landmarksdelineating BA 44 and 45. In particular, in the six participants theyexamined, the anterior ascending ramus of the Sylvian fissure fairlyconsistently demarcated the border of connectivity-parcellated BA 44and 45 (see Fig. 4, page 820 Anwander et al., 2007). In five out of nine ofthe present participants, voxels showing the highest connectivity indexwith Pre-SMA are located posterior to the anterior ascending ramus(younger participants 1, 3, and older participants 4, 5, 8). In all of ourparticipants, voxels with the highest connectivity index with BA 9 arelocated predominantly posterior to the anterior ascending ramus, withonly a few voxels anterior to it. Voxels that have high connectivity indexwith BA 8 are located anterior to the anterior ascending ramus in sevenout of nine participants (younger participants 2, 3, and older participants4, 5, 6, 7, 9), and for the other two these voxels are distributed bothposteriorly and anteriorly to the anterior ascending ramus (youngerparticipant 1, older participant 8). Individual differences in location ofconnectivity-based sub-regions within Broca's area are not surprisinggiven previous histological work on intersubject variability of BA 44 and45 (Amunts et al., 1999). Amunts and colleagues have shown thatalthough in some hemispheres the border between BA 44 and 45 isdefined by the anterior ascending ramus of the Sylvian fissure, in othersit is located close to the diagonal sulcus, or interposed between thesetwo sulci. Apparent disagreements between borders defined byanatomical landmarks, connectivity or histology may reflect factorssuch as brain size, heredity and life experiences.

What does the presently described anterior–posterior connectivitygradient between medial frontal cortex and Broca's area suggest forinvolvement of medial frontal cortex in language? Numerous studieshave shown anterior–posterior functional organization of Broca's area,whereinmore anterior portions corresponding to BA 45 are implicatedin semantic processing, while more posterior portions correspondingto BA 44 are involved in phonological and syntactic processing(Amunts et al., 2004; Hagoort, 2006). Using a verbal fluency task inconjunction with cytoarchitectonic probabilistic maps of BA 44 and 45suggests that the left BA 45 is involved in semantic aspects of languageprocessing, whereas BA 44 is implicated in speech programming(Amunts et al., 2004). Using a similar approach, Binkofski andcolleagues (2000) implicated BA 44 in the mediation of higher-orderforelimb movement control. Support for this functional delineationalso comes from positron emission tomography studies using thecombination of probabilistic cytoarchitectonic maps and functional

activation (Horwitz, et al., 2003) in which activation in BA 44 wasattributed to non-linguistic, laryngeal or limb motor production inspeech or sign language. Thus, based on our findings, one plausibleimplication of the current data is that (a) given the anterior–posteriorconnectivity patterns between Broca's area and medial frontal cortexthen (b) a similar functional gradientmight be found inmedial frontalareas. Piccard and Strick (2001) in their work on SMA and Pre-SMAhave shown that connectivity of a region defines its function. Medialfrontal cortex has been previously implicated to be involved ininitiation of cognitive aspects of spontaneous language (Nielsen andJacobs, 1951a,b; Barris and Schuman, 1953; Luria, 1996; Tijssen et al.,1984). fMRI during phonetic and semantic analysis of aurallypresented stimuli indicated extensive activation of BA 8/9/10 aswell as Pre-SMA during the semantic decision task, whereas activationof the SMA was found only when the subjects were asked to attend tocharacteristics of non-linguistic stimuli (Binder et al., 1997). Lesions tothe SMA/Pre-SMAcomplexhavebeen reported to inducewordfindingdifficulties in speech (Jonas, 1981). Crosson and colleagues (2001)have also demonstrated with fMRI that Pre-SMA is involved ininternally guided word generation. fMRI activation of SMA is alsoseen during volitional expiration, which could suggest involvement ofthis region in control of vocalizations (Ramsay et al., 1993). Letter/character finding is also attributed to Pre-SMA by an fMRI study inwhich participants had to find correct Kanji characters to solve apuzzle (Matsuo et al., 2001). A meta-analysis by Picard and Strick(1996), examining functional activation patterns of SMA and Pre-SMA,suggests that SMA is involved in simple speech tasks including simplerepetition and overpracticed verbal associations, whereas Pre-SMA isinvolved in more complex verbal tasks, such as word production innew conditional associations. Functional organization of medialfrontal cortex does seem to follow to some extent an anterior–posterior gradient similar to that of Broca's area. Data from our ownlaboratory (Crosson et al., 2003) suggests that the story has at least onemore factor. Word generation data from this latter study wererelatively consistent with the anterior-to-posterior gradient for lateralcortices. Specifically, when subjects generated words as exemplars ofcategories (semantic processing), cortex along the inferior frontalsulcus was active; and when subjects generated words that rhymedwith the givenword (phonological processing), premotor cortex alongthe precentral sulcus was active (Broca's region, expanded posterior-ly). However, pre-SMA was active when words were generated fromeither a semantic category or a rhyming cue (lexical but non-semantic). Pre-SMA was not active when nonsense syllables weregenerated from beginning and ending consonant blends. Hence, Pre-SMA was active when generation involved lexical items, which havepre-existing representations in the brain, but not active whengeneration involved non-lexical and non-semantic novel phonologicalitems (nonsense syllables). Clearly the role of lexical and non-lexicalprocessing (as well as semantic, phonological, and syntactic proces-sing) should be taken into account in future investigations of thefunctional relationships linkingmedial frontal cortex and Broca's area.

The current study has some limitations. The participant sample istoo small to identify age-related or gender-related effects. Like otherapplications of magnetic resonance imaging the signals are acquiredthrough volume and time averaging, which sets limits on resolutionand allows for degradation by unwanted head motion. The presentapproach of selecting connectivity candidates based on high empiricalprobability densities of Markov ChainMonte Carlo streamlines cannotrule out overlapping or parallel connections. Unlike high-resolutiontechniques of stains transported through neurons, DW-MRI infersstructure (but it is also noninvasive).

Gender differences in neural architecture have been previouslyinvestigated using fMRI (Shaywitz et al., 1995) as well as DW-MRI(Sullivan et al., 2010; Hsu et al., 2007). Our sample consisting of sevenfemales and twomalesdidnot allowus to investigate genderdifferencesin Broca's area and medial frontal cortex connectivity. Further studies

1236 A. Ford et al. / NeuroImage 52 (2010) 1230–1237

including larger sample sizes and gender-balanced groups of partici-pants are necessary to definitively determine the presence or absence ofgender-related differences in the structural connectivity in question.

Our findings indicate that subjects without SMA representation intheir Broca's area parcellation are younger than those with SMArepresentation. The limited number of participants in our studyprevents us from carrying out a more elaborate formal statisticalcomparison of age-related differences in the connectivity patternsbetween Broca's area and the medial frontal cortex. Future studiesinvestigating changes in the integrity of the white matter connectingBroca's area and the prefrontal cortex to medial frontal cortex wouldbe crucial in determining the underlying causes of the potential age-related differences that could not be addressed extensively in thepresent manuscript.

Inferred structural connectivity was presently guided by anatom-ical landmarks to define Broca's area and regions within medialfrontal cortex. As mentioned above, anatomical landmarks andcytoarchitectonic boundaries often do not agree (Amunts et al.,1999). Functional parcellation provides a more precise definition ofneuroanatomy specific to each participant's functional organization,as demonstrated by Amunts and colleagues (2004). Thus, it is criticalthat future studies should incorporate fMRI and related functionalmodalities to better delineate functional sub-regions within bothBroca's area and medial frontal cortex (Amunts and Zilles, 2001;Horwitz et al., 2003; Eickhoff et al., 2006). The functional informationcould improve our understanding of how and when a particularpathway in the Broca's - medial frontal network is used.

Furthermore, it is crucial to determine whether anterior Broca'sarea co-activates synchronously with anterior medial frontal cortex,and posterior Broca's area co-activates with posterior medial frontalcortex, consistent with the anterior–posterior connectivity patternsdemonstrated in the present study. Partial support for this hypothesiscan be found in work by Kouneiher and colleagues (2009), whoshowed functional connectivity between the posterior lateral pre-frontal cortex, including BA 44/45 and posterior medial frontal cortex(Pre-SMA) on trials evoking contextual motivation. Future studiesimplementing methods similar to those of Kouneiher et al. usinglanguage tasks will provide further insights into functional organiza-tion of the Broca's–medial frontal cortex networks. A furtherchallenge for future studies might be to expand Broca's area intoBroca's region, which includes ventral lateral BA 6 and pars orbitalis(BA 47) (Hagoort, 2006) in addition to pars triangularis (BA 45) andpars opercularis (BA 44), the latter two being traditional Broca's area.

Acknowledgments

This material is based upon work supported by the Office ofResearch and Development, Department of Veterans Affairs (Reha-bilitation R&D Service, Center of Excellence grant #F2182C and SeniorResearch Career Scientist award #B6364L to BC) and by grant #R01DC007387 to BC.

References

Amunts, K., Zilles, K., 2001. Advances in cytoarchitectonic mapping of the humancerebral cortex. Neuroimaging Clin. N. Am. 11 (2), 151–169 vii.

Amunts, K., Schleicher, A., Burgel, U., Mohlberg, H., Uylings, H.B.M., Zilles, K., 1999.Broca's region revisited: cytoarchitecture and intersubject variability. J. Comp.Neurol. 412, 319–341.

Amunts, K., Weiss, P.H., Mohlberg, H., Pieperhoff, P., Gurd, J., Shah, J.N., et al., 2004.Analysis of the neural mechanisms underlying verbal fluency in cytoarchitectoni-cally defined stereotaxic space – The role of Brodmann's areas 44 and 45.Neuroimage 22, 42–56.

Anwander, A., Tittgemeyer, M., von Cramon, D.Y., Friederici, A.D., Knosche, T.R., 2007.Connectivity-based parcellation of Broca's area. Cereb. Cortex 17 (4), 816–825.

Barris, R.W., Schuman, H.R., 1953. Bilateral anterior cingulate gyrus lesions: syndromeof the anterior cingulate gyri. Neurology 3, 44–52.

Basser, P.J., Mattiello, J., LeBihan, D., 1994. MR diffusion tensor spectroscopy andimaging. Biophys. J. 66 (1), 259–267.

Behrens, T.E.J., Johansen-Berg, H., Woolrich, M.W., Smith, S.M., Wheeler-Kingshott, C.A.M., Boulby, P.A., Barker, G.J., Sillery, E.L., Sheehan, K., Ciccarelli, O., Thompson, A.J.,Brady, J.M., Matthews, P.M., 2003a. Non-invasive mapping of connections betweenhuman thalamus and cortex using diffusion imaging. Nat. Neurosci. 6, 750–757.

Behrens, T.E.J., Woolrich, M.W., Jenkinson, M., Johansen-Berg, H., Nunes, R.G., Clare, S.,Matthews, P.M., Brady, J.M., Smith, S.M., 2003b. Characterization and propagation ofuncertainty in diffusion-weighted MR imaging. Magn. Reson. Med. 50, 1077–1088.

Behrens, T.E.J., Johansen-Berg, H., Jbabdi, S., Rushworth, M.F.S., Woolrich, M.W., 2007.Probabilistic diffusion tractography with multiple fibre orientations: What can wegain? NeuroImage 34 (1), 144–155.

Binder, J.R., Frost, J.A., Hammeke, T.A., Cox, R.W., Rao, S.M., Prieto, T., 1997. Human brainlanguage areas identified by functional magnetic resonance imaging. J. Neurosci. 17(1), 353–362.

Binkofski, F., Amunts, K., Stephan, K.M., Posse, S., Schormann, T., Freund, H.J., Zilles, K.,Seitz, R.J., 2000. Broca's region subserves imagery of motion: A combinedcytoarchitectonic and fMRI study. Hum. Brain Mapp. 11 (4), 273–285.

Broca, P., 1861. Remarques sur le siège de la faculté du langage articulé; suivies d'uneobservation d'aphemie. Bull. Soc. Anat. Paris 6, 330–357.

Catani, M., Howard, R.J., Pajevic, S., Jones, D.K., 2002. Virtual in vivo interactivedissection of white matter fasciculi in the human brain. Neuroimage 17 (1),77–94.

Catani, M., Jones, D.K., Ffytche, D.H., 2005. Perisylvian language networks of the humanbrain. Ann. Neurol. 57 (1), 8–16.

Ciccarelli, O., Behrens, T.E.J., Altmann, D.R., Orrell, R.W., Howard, R.S., Johansen-Berg, H.,Miller, D.H., Matthews, P.M., Thompson, A.J., 2006. Probabilistic diffusiontractography: a potential tool to assess the rate of disease progression inamyotrophic lateral sclerosis. Brain 129 (7), 1859–1871.

Courchesne, E., Chisum, H.J., Townsend, J., Cowles, A., Covington, J., Egaas, B., Harwood,M., Hinds, S., Press, G.A., 2000. Normal brain development and aging: quantitativeanalysis at in vivo MR imaging in healthy volunteers. Radiology 216, 672–682.

Cox, R.W., 1996. AFNI software for analysis and visualization of functional magneticresonance neuroimages. Comput. Biomed. Res. 29, 162–173.

Crosson, B., 1999. Subcortical mechanisms in language: lexical–semantic mechanismsand the thalamus. Brain Cogn. 40 (2), 414–438.

Crosson, B., Sadek, J.R., Maron, L., Gokcay, D., Mohr, C.M., Auerbach, E.J., Freeman, A.J.,Leonard, C.M., Briggs, R.W., 2001. Relative shift in activity from medial to lateralfrontal cortex during internally versus externally guided word generation. J. Cogn.Neurosci. 13 (2), 272–283.

Crosson, B., Benefield, H., Cato, M.A., Sadek, J.R., Moore, A.B., et al., 2003. Left andright basal ganglia and frontal activity during language generation: Contributionsto lexical, semantic, and phonological processes. J. Int. Neuropsychol. Soc. 9,1061–1077.

Crosson, B., McGregor, K., Gopinath, K.S., Conway, T.S., Benjamin, M., Chang, Y.L., Moore,A.B., Raymer, A.M., Briggs, R.W., Sherod, M.G., Wierenga, C.E., White, K.D., 2007.Functional MRI of language in aphasia: a review of the literature and themethodological challenges. Neuropsychol. Rev. 17 (2), 157–177.

Dancause, N., Barbay, S., Frost, S.B., Mahnken, J.D., Nudo, R.J., 2007. Interhemisphericconnections of the ventral premotor cortex in a new world primate. J. Comp.Neurol. 505 (6), 701–715.

Eickhoff, S.B., Amunts, K., Mohlberg, H., Zilles, K., 2006. The Human Parietal Operculum.II. Stereotaxic Maps and Correlation with Functional Imaging Results. Cereb. Cortex16 (2), 268–279.

Frey, S., Campbell, J.S.W., Pike, G.B., Petrides, M., 2008. Dissociating the human languagepathways with high angular resolution diffusion fiber tractography. J. Neurosci. 28(45), 11435–11444.

Friedrici, A., 2009. Pathways to language: fiber tracts in the human brain. Trends Cog.Sci. 13 (4), 175–181.

Ghosh, S., Gattera, R., 1995. A comparison of the ipsilateral cortical projections to thedorsal and ventral subdivisions of the macaque premotor cortex. Somatosens. Mot.Res. 12, 359–378.

Glasser, M.F., Rilling, J.K., 2008. DTI tractography of the human brain's languagepathways. Cereb. Cortex 18 (11), 2471–2482.

Hagoort, P., 2006. On Broca, Brain, and Binding. In: Grodzinsky, Y., Amunts, K. (Eds.),Broca's region. Oxford UP, New York.

Heiervang, E., Behrens, T.E.J., Mackay, C.E., Robson, M.D., Johansen-Berg, H., 2006.Between session reproducibility and between subject variability of diffusion MRand tractography measures. Neuroimage 33 (3), 867–877.

Horwitz, B., Rumsey, J.M., Donohue, B.C., 1998. Functional connectivity of the angulargyrus in normal reading and dyslexia. PNAS 95 (15), 8939–8944.

Horwitz, B., Amunts, K., Bhattacharyya, R., Patkin, D., Jeffries, K., Zilles, K., Braun, A.R.,2003. Activation of Broca's area during the production of spoken and signedlanguage: a combined cytoarchitectonic mapping and PET analysis. Neuropsycho-logia 41 (14), 1868–1876.

Hsu, J., Leemans, A., Bai, C., Tsai, Y., Chiu, H., Chen,W., 2007. Gender differences and age-related white matter changes of the human brain: A diffusion tensor imaging study.Neuroimage 39 (2), 566–577.

Jenkinson, M., Smith, S.M., 2001. A global optimisation method for robust affineregistration of brain images. Med. Image Anal. 5 (2), 143–156.

Jonas, S., 1981. The supplementary motor region and speech emission. J. Commun.Disord. 14, 349–373.

Klein, D., Milner, B., Zatorre, R.J., Meyer, E., Evans, A.C., 1995. The neural substratesunderlying word generation: a bilingual functional-imaging study. PNAS 92 (7),2899–2903.

Kouneiher, F., Charron, S., Koechlin, E., 2009. Motivation and cognitive control in thehuman prefrontal cortex. Nat. Neurosci. 12, 939–945.

Lichtheim, L., 1885. On aphasia. Brain 7, 433–484.

1237A. Ford et al. / NeuroImage 52 (2010) 1230–1237

Luria, A.R., 1996. Human brain and psychological processes. Harper & Row, New York.Matsuo, K., Kato, C., Ozawa, F., Takehara, Y., Isoda, H., Isogai, S., Moriya, T., Sakahara, H.,

Okada, T., Nakai, T., 2001. Neuroreport 12 (10), 2227–2230.Nielsen, J.M., Jacobs, L.L., 1951a. Bilateral lesions of the anterior cingulated gyri: report

case. Bull. Los Angeles Neurol. Soc. 16, 231–234.Nielsen, J.M., Jacobs, L.L., 1951b. Bilateral lesions of the anterior cingulate gyri; report of

case. Bull. Los Angeles Neurol. Soc. 16, 231–234.Petrides, M., 1995. A functional organization of the human frontal cortex for mnemonic

processing: evidence from neuroimaging studies. Ann. N. Y. Acad. Sci. 769, 85–96.Petrides, M., 2005. Lateral prefrontal cortex: architectonic and functional organization.

Phil. Trans. R. Soc. B 360 (1456), 781–795.Petrides, M., Pandya, D.N., 1994. Comparative architectonic analysis of the human and

the macaque frontal cortex. In: Boller, F., Grafman, J. (Eds.), Handbook ofNeuropsychology, Vol. 9. Elsevier, Amsterdam, pp. 17–58.

Petrides, M., Pandya, D.N., 2002. Comparative cytoarchitectonic analysis of the humanand the macaque ventrolateral prefrontal cortex and corticocortical connectionpatterns in the monkey. Eur. J. Neurosci. 16, 291–310.

Picard, N., Strick, P.L., 1996. Medial wall motor areas: a review of their location andfunctional activation. Cereb. Cortex 6, 342–353.

Picard, N., Strick, P.L., 2001. Imaging the premotor areas. Current Opinion inNeurobiology 11 (6), 663–672.

Ramsay, S.C., Adams, L., Murphy, K., et al., 1993. Regional cerebral blood flow duringvolitional expiration in man: a comparison with volitional inspiration. J. Physiol.461, 85–101.

Ridderinkhof, K.R., Ullsperger, M., Crone, E.A., Nieuwenhuis, S., 2004. The role of themedial frontal cortex in cognitive control. Science 306 (5695), 443–447.

Rizzolatti, G., Arbib, M.A., 1998. Language within our grasp. Trends Neurosci. 21 (5),188–194.

Rizzolatti, G., Craighero, L., 2004. The mirror-neuron system. Annu. Rev. Neurosci. 27,169–192.

Rushworth, M.F.S., Walton, M.E., Kennerley, S.W., Mannerman, D.M., 2004. Action setsand decisions in the medial frontal cortex. Trends Cogn. Sci. 8 (9), 410–417.

Rushworth, M.F.S., Buckley, M.J., Behrens, T.E.J., Walton, M.E., Bannerman, D.M., 2007.Functional organization of the medial frontal cortex. Curr. Opin. Neurobiol. 17 (2),220–227.

Salat, D.H., Tuch, D.S., Hevelone, N.D., Fischl, B., Corkin, S., Rosas, H.D., Dale, A.M., 2006.Age-related changes in prefrontal white matter measures by diffusion tensorimaging. Ann. N. Y. Acad. Sci. 1064, 37–49.

Shaywitz, B.A., Shaywitz, S.E., Pugh, K.R., Constable, R.T., Skudlarski, P., Fulbright, R.K.,Bronen, R.A., Fletcher, J.M., Shankweiler, D.P., Katz, L., Gores, J.C., 1995. Sex differencesin the functional organization of the brain for language. Nature 373, 607–609.

Smith,M., 2002. Fast robust automated rain extraction.Hum. BrainMapp. 17 (3), 143–155.Sullivan, E.V., Rohlfing, T., Pfefferbaum, A., 2010. Quantitative fiber tracking of the

lateral and interhemispheric white matter systems in normal aging: Relations totimed performance. Neurobiol. Aging 31 (3), 464–481.

Talairach, J., Tournoux, P., 1988. Co-planar stereotaxic atlas of the human brain.Theiler, J., Eubank, S., Longtin, A., Galdrikian, B., Farmer, J.D., 1992. Testing for

nonlinearity in time series: The method of surrogate data. Physica D 58 (77) a.Thieme, Stuttgart, Germany.

Tijssen, C.C., Tavy, D.L.J., Hekster, R.E.M., Bots, G.T.A.M., Endtz, L.J., 1984. Aphasia with aleft frontal interhemispheric hematoma. Neurology 34, 1261–1264.

Wernicke, C., 1874. In: Der aphasische Symptomenkomplex. Breslau, Cohn, Weigert.Westlye, L.T., Walhovd, K.B., Dale, A.M., Bjornerud, A., Due-Tonnessen, P., Engvig, A.,

Grydeland, H., Tammes, C.K., Ostby, Y., Fjell, A.M., in press. Life-span changes in thehuman brain white matter: diffusion tensor imaging and volumetry. CerebralCortex. doi:10.1093/cercor/bhp280.