Elsevier

NeuroImage

Volume 37, Issue 4, 1 October 2007, Pages 1267-1277
NeuroImage

Validation of in vitro probabilistic tractography

https://doi.org/10.1016/j.neuroimage.2007.06.022Get rights and content

Abstract

Diffusion weighted imaging (DWI) and tractography allow the non-invasive study of anatomical brain connectivity. However, a gold standard for validating tractography of complex connections is lacking. Using the porcine brain as a highly gyrated brain model, we quantitatively and qualitatively assessed the anatomical validity and reproducibility of in vitro multi-fiber probabilistic tractography against two invasive tracers: the histochemically detectable biotinylated dextran amine and manganese enhanced magnetic resonance imaging. Post mortem DWI was used to ensure that most of the sources known to degrade the anatomical accuracy of in vivo DWI did not influence the tracking results. We demonstrate that probabilistic tractography reliably detected specific pathways. Moreover, the applied model allowed identification of the limitations that are likely to appear in many of the current tractography methods. Nevertheless, we conclude that DWI tractography can be a precise tool in studying anatomical brain connectivity.

Introduction

Detailed knowledge of the complex anatomical connections in the human central nervous system is pivotal in elucidating and understanding normal and pathological brain function. Our current understanding of the architecture of neuronal circuitry is mainly based upon histological visualization of fiber tracts or neuronal tracing techniques using fluorescent or histochemically detectable tracers (for a review see Kobbert et al., 2000, Oztas, 2003, Zaborszky et al., 2006). Additionally, a new magnetic resonance imaging (MRI) detectable neuronal tracing method has recently been introduced in animal models (Leergaard et al., 2003, Lin et al., 2001, Murayama et al., 2006, Pautler et al., 1998, Saleem et al., 2002). Manganese enhanced MRI (meMRI) exploits the anterograde transport of paramagnetic manganese (Mn2+) and visualizes projection pathways in vivo (Pautler et al., 1998). However, although considered gold standards for assessing brain connectivity, the invasive nature of these techniques restricts their application to experimental animals or post mortem human brain research.

Recently, it has become possible to evaluate anatomical brain connectivity non-invasively by taking advantage of the restricted self-diffusion of water molecules within tissue using diffusion MRI, or diffusion weighted imaging (DWI). Diffusion tensor imaging (DTI) (Basser et al., 1994) and related methods such as q-ball imaging (Tuch, 2004), tensor-mixture models (Alexander, 2005, Assaf et al., 2004, Tuch et al., 2002), PASMRI (Jansons and Alexander, 2003) and spherical deconvolution (Frank, 2002, Tournier et al., 2004) quantify water-diffusion anisotropy in the intra and extra-cellular spaces of the neural environment. Fiber tracking algorithms then use the directions of greatest diffusion as estimates of white matter fiber orientation. Several fiber tracking algorithms have emerged in the last few years that provide reproducible visualizations of three-dimensional fiber tracts (Basser et al., 2000, Jiang et al., 2006, Parker et al., 2002, Parker et al., 2003). One class of these algorithms is probabilistic tractography (Behrens et al., 2003, Parker et al., 2003, Parker and Alexander, 2003, Parker and Alexander, 2005).

Although probabilistic tractography currently holds great promise as a powerful non-invasive connectivity-measurement tool, its accuracy and limitations remain to be evaluated. In particular, a strong correlation between the information about the anatomical connectivity provided by tractgraphy methods and independent anatomical data is still lacking (Behrens et al., 2007, Ciccarelli et al., 2003, Heiervang et al., 2006, Jones et al., 2005). Additionally, the accuracy and precision of tractography results are restricted by limitations on spatial resolution, thermal and physiological noise (for example cardiac pulsation (Skare and Andersson, 2001), respiration-induced signal fluctuations), brain motion and image artefacts (for a review see Le Bihan et al., 2006).

The present study aims to validate the anatomical accuracy of tractography by overcoming most of the known limitations. To this end probabilistic tractography was assessed post mortem in an in vitro environment. Post mortem imaging benefits from high-field (> 3 T) MR scanners and long scanning times, thereby significantly improving the signal-to-noise ratio (SNR) and spatial resolution. Moreover, many of the degrading effects observed in vivo are not present post mortem. Specifically, we evaluated the ability of probabilistic tractography to reconstruct functionally specific pathways detected by invasive neuron tract tracing in the gyrated Göttingen minipig brain.

Section snippets

Tracer injections

Three young male Göttingen minipigs (Ellegaard Göttingen Minipigs A/S, Dalmose, Denmark) were used in the study (age 3 months, weight between 5 and 6 kg). Each pig received a single simultaneous injection of 0.5 μl MnCl2 (0.8 M MnCl2 Tetrahydratem, Sigma-Aldrich, Denmark) and 4.5 μl biotinylated dextran amine (BDA, 10,000 molecular weight, Molecular Probes, Eugene, OR) aimed at the right somatosensory cortex (SC), the right prefrontal cortex (PFC), or the left motor cortex (MC) based on

In vivo tracing

Pathways emanating from the simultaneous injection of two in vivo tracers in the right SC of brain 1, in the right PFC of brain 2 or in the left MC of brain 3 were described anatomically and compared. Injections of BDA into the SC of brain 1 (Fig. 1A) revealed several specific ipsilateral and contralateral corticothalamic, corticostriatal and corticocortical pathways. Dense corticothalamic innervations were observed in the ipsilateral VP nuclei of the thalamus (Figs. 1B, C, F, G). Other fibers

Discussion

In the present study the porcine brain was used as a gyrated brain model in order to quantitatively and qualitatively assess the anatomical validity, consistency and reliability of in vitro probabilistic tractography. Previous validation studies of tractography methods have been based on e.g. simulations (Lazar and Alexander, 2003, Leemans et al., 2005) or groups of human subjects (Behrens et al., 2007, Ciccarelli et al., 2003, Heiervang et al., 2006). Using the porcine brain we demonstrate

Acknowledgments

We thank Matthew G. Liptrot, MSc and Torben Lund, MSc PhD for critical technical input and Birgitte R. Kornum, MSc for assisting during surgery. Thanks to Dr. Klaus Qvortrup and Dr. Maibritt B. Andersen for placing perfusion and stereotaxic equipment at our disposal. Thanks to DVM Nanna Grand and Lars Ellegaard at Ellegaard Minipigs Aps for donating experimental animals. We would like to express our gratitude for the financial support from the Lundbeck Foundation, the Gangsted Foundation, the

References (43)

  • C.P. Lin et al.

    Validation of diffusion tensor magnetic resonance axonal fibre imaging with registered manganese-enhanced optic tracts

    NeuroImage

    (2001)
  • N.M. Lind et al.

    The use of pigs in neuroscience: modeling brain disorders

    Neurosci. Biobehav. Rev.

    (2007)
  • Y. Murayama et al.

    Tracing neural circuits in vivo with Mn-enhanced MRI

    Magn. Reson. Imaging

    (2006)
  • K.S. Saleem et al.

    Magnetic resonance imaging of neuronal connections in the macaque monkey

    Neuron

    (2002)
  • S. Skare et al.

    On the effects of gating in diffusion imaging of the brain using single shot EPI

    Magn. Reson. Imaging

    (2001)
  • J.D. Tournier et al.

    Direct estimation of the fibre orientation density function from diffusion-weighted MRI data using spherical deconvolution

    NeuroImage

    (2004)
  • D.C. Alexander

    Multiple-fibre reconstruction algorithms for diffusion MRI

    Ann. N. Y. Acad. Sci.

    (2005)
  • G.E. Alexander et al.

    Parallel organization of functionally segregated circuits linking basal ganglia and cortex

    Annu. Rev. Neurosci.

    (1986)
  • G.E. Alexander et al.

    Basal ganglia–thalamocortical circuits: parallel substrates for motor, oculomotor, “prefrontal” and “limbic” functions

    Prog. Brain Res.

    (1990)
  • D.C. Alexander et al.

    Detection and modeling of non-Gaussian apparent diffusion coefficient profiles in human brain data

    Magn. Reson. Med.

    (2002)
  • Y. Assaf et al.

    New modeling and experimental framework to characterize hindered and restricted water diffusion in brain white matter

    Magn. Reson. Med.

    (2004)
  • Cited by (188)

    View all citing articles on Scopus
    View full text