Elsevier

NeuroImage

Volume 49, Issue 3, 1 February 2010, Pages 2366-2374
NeuroImage

B-value dependence of DTI quantitation and sensitivity in detecting neural tissue changes

https://doi.org/10.1016/j.neuroimage.2009.10.022Get rights and content

Abstract

Recently, remarkable success has been demonstrated in using MR diffusion tensor imaging (DTI) to characterize white matter. Water diffusion in complex biological tissue microstructure is not a free or Gaussian process but is hindered and restricted, thus contradicting the basic assumption in conventional DTI that diffusion weighted signal decays with b-value in a monoexponential manner. Nevertheless, DTI by far is still the fastest and most robust protocol in routine research and clinical settings. To assess the b-value dependence of DTI indices and evaluate their sensitivities in detecting neural tissues changes, in vivo DTI data acquired from rat brains at postnatal day 13, 21 and 120 with different b-values (0.5–2.5 ms/μm2) and 30 gradient directions were analyzed. Results showed that the mean and directional diffusivities consistently decreased with b-value in both white and gray matters. The sensitivity of axial diffusivity (λ//) in monitoring brain maturation generally decreased with b-value whereas that of radial diffusivity (λ) increased. FA generally varied less with b-value but in a manner dependent of the age and tissue type. Analysis also revealed that the FA sensitivity in detecting specific tissue changes was affected by b-value. These experimental findings confirmed the crucial effect of b-value on quantitative DTI in monitoring neural tissue alterations. They suggested that the choice of b-value in conventional DTI acquisition can be optimized for detecting neural tissue changes but shall depend on the specific tissue type and its changes or pathologies targeted, and caution must be taken in interpreting DTI indices.

Introduction

MR diffusion tensor imaging (DTI) has been shown to provide microstructural information in characterizing tissue microanatomy (Basser, 1995, Basser and Pierpaoli, 1996) that other non-invasive modalities cannot offer. Typical DTI indices, derived from the diffusion tensor as rotationally-invariant parameters, include fractional anisotropy (FA), mean (MD), axial (λ//) and radial (λ) diffusivities. As water diffusion in nerve fibers is anisotropic due to myelination and other inherent axonal structures (Beaulieu, 2002), DTI has demonstrated remarkable success in probing the white matter (WM) integrity, and in describing the orientational neuroarchitecture and connectivity in the central nervous system (CNS). In recent years, DTI has been employed extensively to study the WM associated with both normal physiological and pathophysiological changes, including brain development and aging (Bockhorst et al., 2008, Qiu et al., 2008, Sullivan and Pfefferbaum, 2006, Verma et al., 2005), neurological and psychiatric disorders (Damoiseaux et al., 2008, Kolbe et al., 2009, Roosendaal et al., 2009, Rusch et al., 2007, Song et al., 2004, Sun et al., 2005), brain injuries and tumor (Chan et al., 2009a, Chan et al., 2009b, Kidwell and Wintermark, 2008, Schonberg et al., 2006, Wang et al., 2008, Wang et al., 2009), and cognitive functions (Bucur et al., 2008, Qiu et al., 2008, Teipel et al., 2009).

The 2nd-order three-dimensional diffusion tensor (DT) model assumes that diffusion weighted (DW) signal has a monoexponential dependence on the diffusion-weighting factor (i.e., b-value). In other words, it assumes that water diffusion occurs in a free and unrestricted environment, yielding a Gaussian distribution of water diffusion displacement. However, the complex cellular or axonal microstructures in biological tissues hinder and restrict water molecule diffusion, and lead to restricted or non-Gaussian diffusion. DW signal from biological tissues is thus non-monoexponential with respect to b-value. In addition, because of the anisotropic nature of WM such as the corpus callosum, the extent of diffusion restriction is direction-dependent. Therefore, both directional diffusivities and FA can be b-value dependent, complicating the quantitative and comparative DTI studies.

Despite these fundamental limitations, conventional DTI is still the fastest, relatively robust and accessible protocol for investigation of water diffusion characteristics in neural tissues under routine clinical and research settings. Thus one key question is whether b-value in DTI can be optimized for characterizing the diffusion behaviors and/or their changes associated with specific cellular microstructure or pathology. A number of studies have been reported to optimize b-value for improving the detection of changes involved in brain development (Dudink et al., 2008, Jones et al., 2003), infarction (Meyer et al., 2000, Toyoda et al., 2007) and glioma (Alvarez-Linera et al., 2008, Seo et al., 2008). Several studies also investigated the effect of b-values on MD and FA derived in DTI (Jones and Basser, 2004, Melhem et al., 2000). However, the majority of these studies focused on MD and FA only. It is important to note that FA often cannot distinguish one pathologic condition from another because most pathologies, such as demyelination and axonal damage (Sun et al., 2006), would simply result in FA loss or reduction. Similarly, MD cannot provide the direction along which a specific pathology occurs. Given that the DTI directional analysis has been successfully demonstrated in previous studies for elucidating specific neural tissue pathologies in both humans and animal models (Ewing-Cobbs et al., 2008, Sizonenko et al., 2007, Song et al., 2002, Song et al., 2003, Sun et al., 2006, Trip et al., 2006), it is imperative to investigate the b-value dependence of both FA and directional diffusivities, and its effect on DTI sensitivity in probing neural tissue alterations.

This study aimed to examine the quantitative effect of b-value on DTI indices, and to study the optimal b-value for detecting subtle changes in tissue microstructures. As brain development is accompanied by gradual and local morphological changes in both WM and gray matter (GM) (Bockhorst et al., 2008, Dubois et al., 2006, Huppi and Dubois, 2006), subtle changes in water diffusion characteristics are expected to occur during brain development. Therefore, the postnatal brain maturation in the well-controlled rat model was analyzed in the present study with in vivo DTI that employed various b-values.

Section snippets

Materials and methods

DWIs acquired from the postnatal developing rat brains in a recent study by our group (Cheung et al., 2009) were employed to derive DTI indices for various b-values. These DWIs were originally collected to evaluate the efficacy of diffusion kurtosis imaging. In brief, three groups of normal Sprague–Dawley (SD) rats were scanned. They were postnatal day 13 (P13), 31 (P31) and 120 (P120). Sample size was six for each age group. All experiments were conducted using a Bruker PharmaScan 7T scanner.

DTI index maps computed by DTI using different b-values

Fig. 2 illustrates the typical FA, MD, λ// and λ maps for each age group as computed by DTI model using different two b-value sets (0.0 versus a non-zero b-value) as well as using all six b-values via monoexponential fitting (Monoexp). Note that each type of DTI index maps is displayed in the same grayscale for all three age groups. It can be easily observed that the mean and directional diffusivities generally decreased with b-value. The structural contrasts between WM, GM and cerebrospinal

Discussions

In our previous study, the diffusion kurtosis imaging model that requires DWIs with multiple b-values was shown to offer more sensitive and directionally specific detection of the brain developmental changes in both WM and GM than conventional DTI, including the DTI based on the monoexponential fitting of DWIs with all b-values (Cheung et al., 2009). Despite the inadequacy of DTI model to comprehensively characterize the neural tissue as compared to diffusion kurtosis imaging and other

Conclusions

The effect of b-value on the absolute quantitation of various DTI indices and their sensitivity in detecting tissue microstructural changes during rat brain development was investigated. The results showed that the mean and directional diffusivities consistently decreased with b-value in both white and gray matter. The sensitivity of λ// in monitoring brain maturation generally decreased with b-value whereas that of λ increased. FA generally changed less with b-value but in a manner dependent

Acknowledgment

This work was supported by the Hong Kong Research Grant Council (RGC GRF HKU7808/09M).

References (65)

  • KolbeS. et al.

    Optic nerve diffusion changes and atrophy jointly predict visual dysfunction after optic neuritis

    NeuroImage

    (2009)
  • PfeufferJ. et al.

    Water diffusion in rat brain in vivo as detected at very large b values is multicompartmental

    Magma

    (1999)
  • QiuD. et al.

    Diffusion tensor imaging of normal white matter maturation from late childhood to young adulthood: voxel-wise evaluation of mean diffusivity, fractional anisotropy, radial and axial diffusivities, and correlation with reading development

    NeuroImage

    (2008)
  • RoosendaalS.D. et al.

    Regional DTI differences in multiple sclerosis patients

    NeuroImage

    (2009)
  • RuschN. et al.

    Inferior frontal white matter microstructure and patterns of psychopathology in women with borderline personality disorder and comorbid attention-deficit hyperactivity disorder

    Neuroimage

    (2007)
  • SchonbergT. et al.

    Characterization of displaced white matter by brain tumors using combined DTI and fMRI

    NeuroImage

    (2006)
  • SongS.K. et al.

    Dysmyelination revealed through MRI as increased radial (but unchanged axial) diffusion of water

    Neuroimage

    (2002)
  • SongS.K. et al.

    Diffusion tensor imaging detects and differentiates axon and myelin degeneration in mouse optic nerve after retinal ischemia

    Neuroimage

    (2003)
  • SongS.K. et al.

    Diffusion tensor imaging detects age-dependent white matter changes in a transgenic mouse model with amyloid deposition

    Neurobiol. Dis.

    (2004)
  • SullivanE.V. et al.

    Diffusion tensor imaging and aging

    Neurosci. Biobehav. Rev.

    (2006)
  • SunS.W. et al.

    Detection of age-dependent brain injury in a mouse model of brain amyloidosis associated with Alzheimer's disease using magnetic resonance diffusion tensor imaging

    Exp. Neurol.

    (2005)
  • TripS.A. et al.

    Optic nerve diffusion tensor imaging in optic neuritis

    Neuroimage

    (2006)
  • Alvarez-LineraJ. et al.

    Predicting the histopathological grade of cerebral gliomas using high b value MR DW imaging at 3-tesla

    J. Neuroimaging

    (2008)
  • BasserP.J.

    Inferring microstructural features and the physiological state of tissues from diffusion-weighted images

    NMR Biomed.

    (1995)
  • BeaulieuC.

    The basis of anisotropic water diffusion in the nervous system—a technical review

    NMR Biomed.

    (2002)
  • BennettK.M. et al.

    Characterization of continuously distributed cortical water diffusion rates with a stretched-exponential model

    Magn. Reson. Med.

    (2003)
  • BockhorstK.H. et al.

    Early postnatal development of rat brain: in vivo diffusion tensor imaging

    J. Neurosci. Res.

    (2008)
  • CallaghanP.T.

    Principles of nuclear magnetic resonance microscopy

    (1991)
  • ChanK.C. et al.

    MRI of late microstructural and metabolic alterations in radiation-induced brain injuries

    J. Magn. Reson. Imaging

    (2009)
  • ClarkC.A. et al.

    Water diffusion compartmentation and anisotropy at high b values in the human brain

    Magn. Reson. Med.

    (2000)
  • ClarkC.A. et al.

    Diffusion time dependence of the apparent diffusion tensor in healthy human brain and white matter disease

    Magn. Reson. Med.

    (2001)
  • DamoiseauxJ.S. et al.

    Reduced resting-state brain activity in the “default network” in normal aging

    Cereb. Cortex

    (2008)
  • Cited by (99)

    • Tractography methods and findings in brain tumors and traumatic brain injury

      2021, NeuroImage
      Citation Excerpt :

      These metrics have been used to evaluate the perilesional zone around gliomas, mainly to help distinguish infiltration from edema (Hoefnagels et al., 2014; Provenzale et al., 2004). However, it is noteworthy that FA and ADC can be affected by various factors (Alexander et al., 2001; Hui et al., 2010; Pasternak et al., 2009), and that changes can be nonspecific. Consequently, their clinical values in the diagnostic and prognostic evaluation are limited due to their relatively low specificity and sensitivity.

    • Assessment of peripheral neuropathy in type 2 diabetes by diffusion tensor imaging: A case-control study

      2021, European Journal of Radiology
      Citation Excerpt :

      Our results showed a significant increase in RD in DPN patients and a moderate negative correlation with MCV. According to previous studies on central and peripheral neuropathy, increased RD may reflect demyelinating neuropathy [28–30]. Kronlage et al. [31] proved that RD could be used as a biomarker for myelin integrity by reporting a significant increase in RD of the TN and its negative correlation with MCV in patients with chronic inflammatory demyelinating polyneuropathy (CIDP), a typical demyelinating neuropathy.

    View all citing articles on Scopus
    View full text