Elsevier

NeuroImage

Volume 80, 15 October 2013, Pages 489-504
NeuroImage

Networks of anatomical covariance

https://doi.org/10.1016/j.neuroimage.2013.05.054Get rights and content

Highlights

  • Taxonomy of techniques for studying anatomical covariance (Fig. 1)

  • Applications in neurodevelopment

  • Applications in disease

Abstract

Functional imaging or diffusion-weighted imaging techniques are widely used to understand brain connectivity at the systems level and its relation to normal neurodevelopment, cognition or brain disorders. It is also possible to extract information about brain connectivity from the covariance of morphological metrics derived from anatomical MRI. These covariance patterns may arise from genetic influences on normal development and aging, from mutual trophic reinforcement as well as from experience-related plasticity. This review describes the basic methodological strategies, the biological basis of the observed covariance as well as applications in normal brain and brain disease before a final review of future prospects for the technique.

Introduction

In the last decade, there has been an explosion of interest in the study of brain “connectivity”, the systems-level relationship among brain regions (Bullmore and Sporns, 2009, Bullmore and Sporns, 2012, Guye et al., 2010, He and Evans, 2010, Sporns, 2012, Wen et al., 2011, Xia and He, 2011). Many studies focus upon functional connectivity, using correlation analysis to identify putative regional connections, either with EEG/MEG (Micheloyannis et al., 2006, Stam, 2004, Stam et al., 2007) or fMRI (Achard and Bullmore, 2007, Achard et al., 2006, Bassett and Bullmore, 2006, Bellec et al., 2010, Biswal et al., 2010, Eguiluz et al., 2005, Greicius et al., 2003, Greicius et al., 2004, He et al., 2009a, Honey et al., 2007, Honey et al., 2009, Salvador et al., 2005a, Salvador et al., 2005b, Smith et al., 2009, Vértes et al., 2012, Zuo et al., 2010). There has been a similar boom in studies of structural connectivity. Many groups employ diffusion-weighted imaging (DWI) and tractography to examine structural connectivity as a white matter (WM) phenomenon (Behrens et al., 2007, Gong et al., 2009a, Gong et al., 2011, Hofer and Frahm, 2006, Iturria-Medina, 2013, Jbabdi et al., 2007, Le Bihan and Johansen-Berg, 2012, Raj and Chen, 2011, Smith et al., 2006). In contrast to DWI approaches, there has been a growing interest in correlation of cortical grey matter (GM) morphology across the brain (Alexander-Bloch et al., 2013b, Chen et al., 2008, Chen et al., 2011a, Chen et al., 2011b, Gong et al., 2009b, Gong et al., 2011, Gong et al., 2012, He and Evans, 2010, He et al., 2007, He et al., 2008, He et al., 2009a, He et al., 2009b, Khundrakpam et al., in press, Lerch et al., 2006, Lo et al., 2011, Mechelli et al., 2005, Raj et al., 2012, Ramnani et al., 2004, Raznahan et al., 2011a, Reid and Evans, in press, Reid et al., 2010, Seeley et al., 2009, Zhou et al., 2012, Zielinski et al., 2010). Human cortical morphology varies dramatically across individuals (Kennedy et al., 1998) and genetic influence over brain morphology has been repeatedly demonstrated (Brans et al., 2010, Lenroot et al., 2009, Peper et al., 2009, Shaw et al., 2007a, Shaw et al., 2007b, Shaw et al., 2009, van Soelen et al., 2012a, van Soelen et al., 2012b, Wright et al., 2002, Yoon et al., 2010, Yoon et al., 2011, Yoon et al., 2012). However, while strong anatomical covariance has been observed in post-mortem studies (Andrews et al., 1997), it is not clear to what extent this covariance is a result of genetic influence (Chen et al., 2011a, Chen et al., 2011b, Chen et al., 2012, Schmitt et al., 2008, Schmitt et al., 2009, Schmitt et al., 2010), normal development and aging (Chen et al., 2011a, Chen et al., 2011b, Raz et al., 2005), mutual trophic influences (Aid et al., 2007, Ferrer et al., 1995) or experience-related plasticity (Bermudez et al., 2009, Draganski et al., 2004, Driemeyer et al., 2008, Haier et al., 2009, Hyde et al., 2009a, Hyde et al., 2009b, Maguire et al., 2000, Maguire et al., 2003, Maguire et al., 2006, Schneider et al., 2002). This anatomical covariance structure is amenable to network analysis in a manner analogous to functional network analysis. As will be developed below, the graph properties of networks derived from anatomical covariance change with normal development and aging as well as with disease type and severity. Structural network analysis therefore holds great promise in many areas of neuroscience. However, a number of outstanding issues remain to be addressed. Graph theoretical models, drawn from social science and applied with little adaptation to neuroscience questions (Bassett and Bullmore, 2006, Sporns et al., 2000, Sporns et al., 2005), do not adequately reflect the underlying neuroanatomy in representing graph “nodes” and “edges”. We need to better understand the impact of regional segmentation and connectivity weighting on network metrics. Furthermore, the microstructural basis of the macroscopically-observed morphological correlation is still unclear. In a recent review of the evolving landscape of human cortical connectivity, Mesulam (Mesulam, 2012) stated that “to enlarge the currently limited data set on structural connectivity is of considerable importance for conducting biologically more valid explorations of large-scale neurocognitive networks. This challenging goal will require histological laboratory investigations of the human brain to resume their former prominence and to play an increasingly more substantial role in brain mapping research”. This review provides a review of the history, current status and future of GM covariance analysis and its applications in brain research.

Section snippets

Methodology

In contrast to studies of functional connectivity that explore the correlation of regional fMRI or EEG/MEG signals across time within an individual, many studies of anatomical covariance examine the correlation among regional structural measures across subjects. It is important to note that a significant inter-regional morphometric correlation does not demand the existence of a direct fibre connection between the regions. As with functional correlation, anatomical correlation can arise from an

Genetics

There is now an extensive literature regarding genetic influences upon the coordinated growth of spatially separated regions (Tost et al., 2012 and references therein). An early study (Pezawas et al., 2005) revealed the impact of a 5-HTTLPR polymorphism for the serotonin transporter gene. Carriers have increased anxiety-related temperamental traits, increased amygdala reactivity and elevated risk of depression. They also showed reduced gray matter volume, as well as function uncoupling, for

Experience-related changes

At the time of writing, there has been no explicit study of changes in anatomical covariance as a consequence of experience or learning. Numerous studies have reported focal changes in morphology that may arise from experience or training (Bermudez et al., 2009, Lv et al., 2008, Maguire et al., 2000, Maguire et al., 2003, Maguire et al., 2006, Schneider et al., 2002) although there is always the caveat in a cross-sectional study that performance is a consequence of morphology rather than the

Comparison with other forms of brain connectivity

A central question for anatomical covariance analysis is the extent to which the observed GM correlation network compares with that derived diffusion-weighted imaging (DWI) of white matter pathways or from functional correlation of fMRI time courses. DWI expanded rapidly from its earliest origins (Basser et al., 1994, Basser et al., 2000, Jones et al., 1999, Le Bihan et al., 1986) with many advances in acquisition, e.g. HARDI (Tuch et al., 2002) and analysis (Behrens et al., 2007,

Development

There is now a massive literature on neuroanatomical maturation, capturing focal changes in cortical morphology with age (e.g. Giedd et al., 1999, Raznahan et al., 2011b, Shaw et al., 2008, Sowell et al., 2004) or the neural correlates of behavioural variables, such as IQ (Karama et al., 2009, Karama et al., 2011, Shaw et al., 2006). However, that literature generally does not deal with the issue of anatomical covariance in pediatric populations.

This being said, structural covariance analysis

Applications in disease

Anatomical covariance strategies have the potential to reveal more about the etiology of neuropathology than a focal approach since they better reflect the distributed nature of neural activity that underlies clinical assessments of disease (Bullmore and Sporns, 2009, Guye et al., 2010, He et al., 2009b, Lo et al., 2011, Reid and Evans, in press).

The Davatzikos group (Fan et al., 2008, Misra et al., 2009) have used high-dimensional classification of GM density to examine the covariance patterns

Future

The study of networks of anatomical covariance has come of age in the last few years. Many groups now employ variations on the general theme to examine the neural correlates of neurodevelopment, learning, normal aging and brain. However, there remain obvious challenges in the current state of the art. Graph theoretical models, drawn from social science and applied with little adaptation to neuroscience questions, do not adequately reflect the underlying neuroanatomy in representing graph

References (254)

  • M. Bastiani et al.

    Human cortical connectome reconstruction from diffusion weighted MRI: the effect of tractography algorithm

    NeuroImage

    (2012)
  • T.E.J. Behrens et al.

    Probabilistic diffusion tractography with multiple fibre orientations: what can we gain?

    NeuroImage

    (2007)
  • P. Bellec et al.

    Multi-level bootstrap analysis of stable clusters in resting state fMRI

    NeuroImage

    (2010)
  • K.L. Bergfield et al.

    Age-related networks of regional covariance in MRI gray matter: reproducible multivariate patterns in healthy aging

    NeuroImage

    (2010)
  • B.C. Bernhardt et al.

    Mapping limbic network organization in temporal lobe epilepsy using morphometric correlations: Insights on the relation between mesiotemporal connectivity and cortical atrophy

    NeuroImage

    (2008)
  • B.C. Bernhardt et al.

    Thalamo–cortical network pathology in idiopathic generalized epilepsy: Insights from MRI-based morphometric correlation analysis

    NeuroImage

    (2009)
  • H. Braak et al.

    Morphological criteria for the recognition of Alzheimer's disease and the distribution pattern of cortical changes related to this disorder

    Neurobiol. Aging

    (1994)
  • A.M. Brickman et al.

    Structural MRI covariance patterns associated with normal aging and neuropsychological functioning

    Neurobiol. Aging

    (2007)
  • E.T. Bullmore et al.

    Does dysplasia cause anatomical dysconnectivity in schizophrenia?

    Schizophr. Res.

    (1998)
  • E.J. Canales-Rodríguez et al.

    Diffusion orientation transform revisited

    NeuroImage

    (2010)
  • E.J. Canales-Rodríguez et al.

    Deconvolution in diffusion spectrum imaging

    NeuroImage

    (2010)
  • E.J. Canales-Rodríguez et al.

    Statistical analysis of brain tissue images in the wavelet domain: wavelet-based morphometry

    NeuroImage

    (2013)
  • C.H. Chen et al.

    Genetic influences on cortical regionalization in the human brain

    Neuron

    (2011)
  • D.B. Chklovskii et al.

    Wiring optimization in cortical circuits

    Neuron

    (2002)
  • Y. Cho et al.

    Individual subject classification for Alzheimer's disease based on incremental learning using a spatial frequency representation of cortical thickness data

    NeuroImage

    (2012)
  • M.K. Chung et al.

    A unified statistical approach to deformation-based morphometry

    NeuroImage

    (2001)
  • G. Collin et al.

    Disturbed grey matter coupling in schizophrenia

    Eur. Neuropsychopharmacol.

    (2013)
  • J.J. Crofts et al.

    Network analysis detects changes in the contralesional hemisphere following stroke

    NeuroImage

    (2011)
  • A.M. Dale et al.

    Cortical surface-based analysis. I. Segmentation and surface reconstruction

    NeuroImage

    (1999)
  • J.M. Duarte-Carvajalino et al.

    Hierarchical topological network analysis of anatomical human brain connectivity and differences related to sex and kinship

    NeuroImage

    (2012)
  • A.C. Evans et al.

    NIH MRI study of normal brain development

    NeuroImage

    (2006)
  • A.C. Evans et al.

    Brain templates and atlases

    NeuroImage

    (2012)
  • Y. Fan et al.

    Spatial patterns of brain atrophy in MCI patients, identified via high-dimensional pattern classification, predict subsequent cognitive decline

    NeuroImage

    (2008)
  • Y. Fan et al.

    Brain anatomical networks in early human brain development

    NeuroImage

    (2011)
  • I. Ferrer et al.

    Transforming growth factor-alpha immunoreactivity in the developing adult brain

    Neuroscience

    (1995)
  • B. Fischl et al.

    Cortical surface-based analysis. II: Inflation, flattening, and a surface-based coordinate system

    NeuroImage

    (1999)
  • G. Gong et al.

    Convergence and divergence of cortical thickness correlations with anatomical connections in the human cerebral cortex

    NeuroImage

    (2012)
  • S. Achard et al.

    Efficiency and cost of economical brain functional networks

    PLoS Comput. Biol.

    (2007)
  • S. Achard et al.

    A resilient, low-frequency, small-world human brain functional network with highly connected association cortical hubs

    J. Neurosci.

    (2006)
  • T. Aid et al.

    Mouse and rat BDNF gene structure and expression revisited

    J. Neurosci. Res.

    (2007)
  • R. Albert et al.

    Error and attack tolerance of complex networks

    Nature

    (2000)
  • G.E. Alexander et al.

    Regional network of magnetic resonance imaging gray matter volume in healthy aging

    Neuroreport

    (2006)
  • A.F. Alexander-Bloch et al.

    The anatomical distance of functional connections predicts brain network topology in health and schizophrenia

    Cereb. Cortex

    (2013)
  • A.F. Alexander-Bloch et al.

    The convergence of maturational change and structural covariance in human cortical networks

    J. Neurosci.

    (2013)
  • T.J. Andrews et al.

    Correlated size variations in human visual cortex, lateral geniculate nucleus, and optic tract

    J. Neurosci.

    (1997)
  • Y. Assaf et al.

    AxCaliber: a method for measuring axon diameter distribution from diffusion MRI

    Magn. Reson. Med.

    (2008)
  • D. Barazany et al.

    In vivo measurement of axon diameter distribution in the corpus callosum of rat brain

    Brain

    (2009)
  • P.J. Basser et al.

    In vivo fiber tractography using DT-MRI data

    Magn. Reson. Med.

    (2000)
  • D.S. Bassett et al.

    Small-world brain networks

    Neuroscientist

    (2006)
  • D.S. Bassett et al.

    Hierarchical organization of human cortical networks in health and schizophrenia

    J. Neurosci.

    (2008)
  • Cited by (323)

    • Resolving heterogeneity in Alzheimer's disease based on individualized structural covariance network

      2024, Progress in Neuro-Psychopharmacology and Biological Psychiatry
    • Gray matter morphological abnormities are constrained by normal structural covariance network in OCD

      2024, Progress in Neuro-Psychopharmacology and Biological Psychiatry
    View all citing articles on Scopus
    View full text