Elsevier

NeuroImage

Volume 55, Issue 4, 15 April 2011, Pages 1657-1664
NeuroImage

Stochastic tractography study of Inferior Frontal Gyrus anatomical connectivity in schizophrenia

https://doi.org/10.1016/j.neuroimage.2011.01.047Get rights and content

Abstract

Background

Abnormalities within language-related anatomical structures have been associated with clinical symptoms and with language and memory deficits in schizophrenia. Recent studies suggest disruptions in functional connectivity within the Inferior Frontal Gyrus (IFG) network in schizophrenia. However, due to technical challenges, anatomical connectivity abnormalities within this network and their involvement in clinical and cognitive deficits have not been studied.

Material and methods

Diffusion and anatomical scans were obtained from 23 chronic schizophrenia patients and 23 matched controls. The IFG was automatically segmented, and its white matter connections extracted and measured with newly-developed stochastic tractography tools. Correlations between anatomical structures and measures of semantic processing were also performed.

Results

White Matter connections between the IFG and posterior brain regions followed two distinct pathways: dorsal and ventral. Both demonstrated left lateralization, but ventral pathway abnormalities were only found in schizophrenia. IFG volumes also showed left lateralization and abnormalities in schizophrenia. Further, despite similar laterality and abnormality patterns, IFG volumes and white matter connectivity were not correlated with each other in either group. Interestingly, measures of semantic processing correlated with white matter connectivity in schizophrenia and with gray matter volumes in controls. Finally, hallucinations were best predicted by both gray matter and white matter measures together.

Conclusions

Our results suggest abnormalities within the ventral IFG network in schizophrenia, with white matter abnormalities better predicting semantic deficits. The lack of a statistical relationship between coexisting gray and white matter deficits might suggest their different origin and the necessity for a multimodal approach in future schizophrenia studies.

Research highlights

► IFG volume and its connectivity left lateralization in both groups. ► No correlation between IFG volumes and white matter connectivity. ► Ventral language white matter pathway abnormalities in schizophrenia. ► Semantic processing related to connectivity only in schizophrenia. ► Hallucinations predicted by both gray and white matter measures together.

Introduction

Fronto–temporal connectivity abnormalities have long been postulated as implicated in the etiology of schizophrenia (Kraepelin, 1919/1971, Wernicke, 1906). These ideas have been further fueled by relatively recent studies (McGuire and Frith, 1996, Weinberger et al., 1992) which have framed the disconnectivity hypothesis as explicitly functional in nature, reflecting disease-related abnormalities in communication and coordination of networks of neural regions (Friston and Frith, 1995). As further support for this hypothesis, fMRI studies in schizophrenia point to a reduction in correlated activation in frontal-temporal networks (e.g., Kubicki et al., 2003, Meyer-Lindenberg et al., 2005, Jeong et al., 2009, Lui et al., 2009). More recent studies, fueled in part by post mortem and genetic evidence of myelination abnormalities (e.g., Davis et al., 2003, Segal et al., 2007), are focused on anatomical disconnectivity in schizophrenia. Arguably the strongest evidence in favor of anatomical disconnectivity comes from studies that have used Diffusion Tensor Imaging (DTI), a method sensitive to white matter fiber tract integrity. Since the development of (DTI) (Basser et al., 1994), this method has become the tool of choice in studying brain connectivity in healthy subjects and its disruptions in various neuropsychological diseases.

The overwhelming advantages of DTI over other methods used to study connectivity (such as ERP, MEG, and animal post mortem tractography methods) include its wide availability, noninvasiveness, high spatial resolution, and the fact that it provides visualization capabilities heretofore not possible. Several studies, to date, have investigated white matter connections (tracts) between frontal and temporal cortices (including the Uncinate Fasciculus (UF), Cingulum Bundle (CB); Fornix, Inferior Occipito–Frontal Fasciculus (IOFF) and Arcuate Fasciculus (AF); for recent review of schizophrenia DTI literature, see Kubicki et al., 2007) in schizophrenia, but despite the potential statistical power of these experiments, results still remain inconclusive. Clinical and cognitive correlates of these connections also remain elusive.

Among the anatomical connections of interest in schizophrenia are those relevant to language processing. Of note here, cognitive deficits observed in schizophrenia are frequently associated with language and verbal memory impairments (e.g., Saykin et al., 1991, Saykin et al., 1994, Adams et al., 1993, Kareken et al., 1996, Nestor et al., 1997). Functional studies report abnormal activation in both frontal and temporal regions in schizophrenics during tasks that utilize linguistic stimuli (Heckers et al., 1998, Ragland et al., 2001, Stevens et al., 1998, Yurgelun-Todd et al., 1996). Additionally, hallmark clinical symptoms of schizophrenia, namely hallucinations, delusions, and thought disorder, are also strongly related to language production and semantic processing (Ceccherini-Nelli et al., 2007, Ford et al., 2002, Han et al., 2007, Lawrie et al., 2002), as well as linked with anatomical abnormalities within both frontal and temporal lobes. Based on these findings, several theories postulate language and its dysfunction as central to schizophrenia psychopathology (e.g., Crow, 1997, Ford et al., 2002). It has been demonstrated that distant brain regions do not work independently, but form functional “networks”, where several different regions communicate while performing specific tasks. These networks are interconnected usually by more than one connection, and disruption of one specific tract does not necessary lead to functional deficits, since the other connection might be able to play a compensatory role. The fronto-temporal connections involved in language/semantic network are no exception. Recent functional and anatomical studies suggest that two distinct “processing streams” provide communication between two regions that are crucial to speech processing (posterior temporal region of Wernicke and inferior frontal region of Broca). These processing streams, called ventral and dorsal (Duffau, 2008, Haroon et al., 2006, Saur et al., 2008) (see also Fig. 1) both contain direct and indirect white matter tracts (AF being the direct dorsal connection, and IOFF forming the direct ventral connection). In addition to direct connections, several studies point to indirect connections between those two regions, that would involve intermediate connections with Geshwind area (Catani et al., 2005, Catani et al., 2007), as well as Uncinate and Inferior/Medial Longitudinal Fasciculi (Mandonnet et al., 2007). Thus anatomical evaluation of the integrity of language/semantic network and its clinical and cognitive correlates should involve modeling and measuring both gray matter volumes as well as white matter connections. When successful, such studies should help to reveal specific roles played by gray and white matter pathology, as well as their synergic effect on clinical and cognitive schizophrenia symptomatology. Thus far, however, studies that combine measurements of gray and white matter integrity within the functionally homogenous regions of the brain are rare likely due to the technical challenges.

More specifically, thus far, most DTI studies investigating specific white matter connections in the brain use the streamline (otherwise called principal diffusion direction) tractography method. This popular method estimates tracts by following the direction of maximal water diffusion of the white matter voxels. Unfortunately streamline tractography does not provide information about the confidence regarding the estimated fiber bundles, thus uncertainty of the generated tracks caused by increased imaging noise (such as diffusion signal within the gray matter) or complex fiber configurations (such as fiber crossings) are not taken into account. Streamline tractography is thus not an optimal tool for studying connectivity between gray matter regions. The stochastic tractography method, introduced recently (Björnemo et al., 2002), is a Bayesian approach that addresses the aforementioned shortcomings of streamline tractography by performing tractography under a probabilistic framework that accounts for uncertainty in diffusion tensor fields. This method uses probabilistic models of imaging noise and fiber architecture to infer the underlying fiber configuration, and, since it explicitly models uncertainty, and does not use any stopping criteria for generating tracts, the method is not limited to generating tracts in regions of low uncertainty/low FA. Consequently, stochastic tractography can track through fiber crossings and be initiated in gray matter, which makes it a perfect tool to model and measure the anatomy of specific functional networks. In addition, certainty of white matter connection measured along the tract can provide additional information regarding the strength of anatomical connectivity between two ROIs (as proposed also recently by Kreher et al., 2008).

In this study, we use stochastic tractography in order to model the direct connections involved in fronto-temporal language/semantic network as well as to detect abnormalities in its integrity and connectivity in schizophrenia. Since such a model should include both gray matter regions as well as white matter connections, we measure gray matter volumes of the inferior frontal and superior temporal gyri, connections between those two regions (white matter), relationships (correlations) between gray and white matter measurements, and their association with clinical symptoms as well as neuropsychological tests related to semantic processing.

Section snippets

Subjects

Twenty three patients with chronic schizophrenia were recruited from in-patient, day treatment, out-patient, and foster care programs at the VA Boston Healthcare System, Brockton, MA. DSM-IV diagnoses were based on SCID-P interviews, and information from patient medical records. Twenty three comparison subjects (group-matched to patients on age, sex, handedness (Edinburgh Handedness Inventory), and parental social economic-status (PSES)) were recruited through advertisements in local

Demographic data

Groups did not differ in age at the time of scan (P(1.44) = 0.236, t =  1.20), in handedness (P(1.44) = 0.99, t =  0.002) nor in gender (all males). Schizophrenics, similar to our previous investigations, had fewer years of education and lower SES than controls, but had comparable PSES with controls (P(1.44) = 0.48, t =  0.71). In addition, premorbid intelligence, tested using WRAT reading raw scores, showed no group differences (P(1.44) = 0.21, t = 1.28). Schizophrenics had a mean duration of illness of 16.2 

Discussion

Our results demonstrate reduced integrity in several anatomical structures involved in semantic processing in schizophrenia. These structures include left IFG gray matter, as well as white matter connections traveling through the temporal stem and connecting the IFG and STG. Interestingly, while gray matter integrity correlated with several cognitive measures of semantic processing in healthy subjects, the same cognitive measures in schizophrenics correlated with white matter integrity rather

Conclusions

Using high resolution DTI data and stochastic tractography techniques, we were able to map anatomical connections between the IFG and STG, and to measure them in chronic schizophrenia. Our results suggest abnormalities within the ventral processing stream in schizophrenia, further suggesting that, while volume of gray matter structures may predict semantic abilities, white matter abnormalities better predict cognitive deficits in this domain in schizophrenia. Even though gray and white matter

Acknowledgments

We gratefully acknowledge the support of the National Institute of Health (K05 MH070047 and R01 MH 50740 to MES, R01 MH 40799 to RWM, R01 MH 082918 to SB and R01MH 074794 to CFW, P50MH 080272 to RWM, MES), the Department of Veterans Affairs Merit Awards (MES, RWM), the VA Schizophrenia Center Grant (RWM/MES), and a Middleton Award (RWM) from the Department of Veterans Affairs. This work is also part of the National Alliance for Medical Image Computing (NAMIC), funded by the National Institutes

References (66)

  • M. Kubicki et al.

    Voxel-based morphometric analysis of gray matter in first episode schizophrenia

    Neuroimage

    (2002)
  • M. Kubicki et al.

    Cingulate fasciculus integrity disruption in schizophrenia: a magnetic resonance diffusion tensor imaging study

    Biol. Psychiatry

    (2003)
  • M. Kubicki et al.

    DTI and MTR abnormalities in schizophrenia: analysis of white matter integrity

    Neuroimage

    (2005)
  • M. Kubicki et al.

    A review of diffusion tensor imaging studies in schizophrenia

    J. Psychiatr. Res.

    (2007)
  • S.M. Lawrie et al.

    Reduced frontotemporal functional connectivity in schizophrenia associated with auditory hallucinations

    Biol. Psychiatry

    (2002)
  • R.W. McCarley et al.

    MRI anatomy of schizophrenia

    Biol. Psychiatry

    (1999)
  • R.A. Morey et al.

    A comparison of automated segmentation and manual tracing for quantifying hippocampal and amygdala volumes

    Neuroimage

    (2009)
  • O.R. Phillips et al.

    Fiber tractography reveals disruption of temporal lobe white matter tracts in schizophrenia

    Schizophr. Res.

    (2009)
  • H.W. Powell et al.

    Hemispheric asymmetries in language-related pathways: a combined functional MRI and tractography study

    Neuroimage

    (2006)
  • M.E. Shenton et al.

    A review of MRI findings in schizophrenia

    Schizophr. Res.

    (2001)
  • K. Vogeley et al.

    Compartmental volumetry of the superior temporal gyrus reveals sex differences in schizophrenia — a post-mortem study

    Schizophr. Res.

    (1998)
  • H. Yamasue et al.

    Localized volume reduction in prefrontal, temporolimbic, and paralimbic regions in schizophrenia: an MRI parcellation study

    Psychiatry Res.

    (2004)
  • A.L. Alexander et al.

    Elimination of eddy current artifacts in diffusion-weighted echo-planar images: the use of bipolar gradients

    Magn. Reson. Med.

    (1997)
  • M. Ashtari et al.

    Disruption of white matter integrity in the inferior longitudinal fasciculus in adolescents with schizophrenia as revealed by fiber tractography

    Arch. Gen. Psychiatry

    (2007)
  • F.M. Benes

    Why does psychosis develop during adolescence and early adulthood?

    Curr. Opin. Psychiatry

    (2003)
  • F.M. Benes et al.

    Myelination of a key relay zone in the hippocampal formation occurs in the human brain during childhood, adolescence, and adulthood

    Arch. Gen. Psychiatry

    (1994)
  • J.R. Binder et al.

    Human brain language areas identified by functional magnetic resonance imaging

    J. Neurosci.

    (1997)
  • M. Björnemo et al.

    Regularized stochastic white matter tractography using diffusion tensor MRI

  • M.S. Buchsbaum et al.

    Diffusion tensor imaging of frontal lobe white matter tracts in schizophrenia

    Ann. Gen. Psychiatry

    (2006)
  • M. Catani et al.

    Perisylvian Language Networks of the Human Brain

    Ann. Neurol.

    (2005)
  • M. Catani et al.

    Symmetries in Human Brain Language Pathways Correlate with Verbal Recall

    PNAS

    (2007)
  • T.J. Crow

    Schizophrenia as failure of hemispheric dominance for language

    Trends Neurosci.

    (1997)
  • T.J. Crow

    Nuclear schizophrenic symptoms as a window on the relationship between thought and speech

    Br. J. Psychiatry

    (1998)
  • Cited by (41)

    • Models of Network Spread and Network Degeneration in Brain Disorders

      2018, Biological Psychiatry: Cognitive Neuroscience and Neuroimaging
      Citation Excerpt :

      Functional changes, such as cortical atrophy and electroencephalography, are accompanied by extensive structural damage (87), which lends support to dynamics of networks hypotheses. Diffusion MRI studies report altered white matter integrity in SZ (88–91), and graph theory analysis shows widespread disturbances in structural organization, as reflected by metrics such as path length, centrality, efficiency, and clustering coefficient (8–10,92–94). As in any developmental disorder, many of the network-level alterations in SZ have a strong age effect.

    View all citing articles on Scopus
    View full text