Elsevier

NeuroImage

Volume 19, Issue 3, July 2003, Pages 545-554
NeuroImage

Regular article
DTI mapping of human brain connectivity: statistical fibre tracking and virtual dissection

https://doi.org/10.1016/S1053-8119(03)00142-3Get rights and content

Abstract

Several approaches have been used to trace axonal trajectories from diffusion MRI data. If such techniques were first developed in a deterministic framework reducing the diffusion information to one single main direction, more recent approaches emerged that were statistical in nature and that took into account the whole diffusion information. Based on diffusion tensor MRI data coming from normal brains, this paper presents how brain connectivity could be modelled globally by means of a random walk algorithm. The mass of connections thus generated was then virtually dissected to uncover different tracts. Corticospinal, corticobulbar, and corticothalamic tracts, the corpus callosum, the limbic system, several cortical association bundles, the cerebellar peduncles, and the medial lemniscus were all investigated. The results were then displayed in the form of an in vivo brain connectivity atlas. The connectivity pattern and the individual fibre tracts were then compared to known anatomical data; a good matching was found.

Introduction

Recent developments in diffusion MRI have put this imaging modality to the forefront of interest among the neuroscientific community. The success of diffusion imaging is related to the fact that during their random, diffusion-driven displacements, water molecules probe tissue structure at a microscopic scale well beyond the usual imaging resolution (LeBihan et al., 2001). It has been shown that, in the brain, ordered axonal structure, cell membrane, and myelin sheath strongly influence water diffusion Beaulieu and Allen 1994, Beaulieu 2002 and that there is a direct link between water diffusion and axonal orientation and integrity Coremans et al 1994, Wieshmann et al 1999. In fact, when diffusion tensor (DT) imaging is performed within a compact tract with parallel running axonal trajectories like the corticospinal tract, the DT is strongly anisotropic and its principal eigenvector corresponds to the direction of the fibre tract.

These observations were used by several researchers to develop fibre tracking algorithms that all have the same aim: inferring from a DT field the axonal or at least bundles of fibres trajectories—the diameter of an axon being well beyond the resolution of a current MRI scan. Impressive results have been achieved and a wide spectrum of applications is foreseen. A better understanding of diffusion properties in many brain-related diseases, e.g., multiple sclerosis Maldjian 2001, Filippi et al 2001, dyslexia (Klingberg et al., 2000), Alzheimer’s disease Rose et al 2000, Bozzali et al 2002, schizophrenia Lim et al 1999, Foong et al 2000, brain tumours Field et al 2002, Mori et al 2002a, periventricular leukomalacia (Hoon et al., 2002), as well as spinal cord injury (Mamata et al., 2002) should benefit from those developments. The understanding of normal brain function needs not only the description of activated cortical areas, like that provided by fMRI, but also a detailed description of the underlying neuronal circuitry.

Most of the algorithms used to infer bundles of fibres from DT imaging are based on a discrete resolution of the integral curves of the vector field corresponding to the reduction of the diffusion tensor to its largest eigenvector Conturo et al 1999, Mori et al 1999, Jones et al 1999, Basser et al 2000, Tench et al 2002. As opposed to those deterministic integral path approaches, this work investigates brain circuitry with a statistical random-walk-based algorithm.

Section snippets

Mri data acquisition

The images used for this study were acquired with a 1.5 T clinical MRI scanner (Magnetom Symphony; Siemens, Erlangen, Germany). The data were produced with a diffusion-weighted single-shot EPI sequence using the standard Siemens Diffusion Tensor Imaging Package for Symphony. We acquired 44 axial slices in a 128 by 128 matrix size covering the whole brain. The voxel size was 1.64 by 1.64 mm with a slice thickness of 3.00 mm without gap. Timing parameters were a TR of 1000 s and a TE of 89 s.

Results

The evaluation phase of this research work was performed on two healthy volunteers. It allowed assessing the capacity of our tracking algorithm to infer axonal connectivity by comparing the results to postmortem-based neuroanatomical knowledge, using the Nieuwenhuys et al. (1988) atlas. We applied the above-mentioned methodology in order to perform a virtual dissection of several well-known anatomical systems. This collection of connectivity maps can also be considered to be the beginning of an

Discussion

We developed a new approach to map brain connectivity that was statistical in nature and based on a global approach toward fibre tracking. This lead to the identification of several fibre tracts (Fig. 6) that all showed accurate correlation with current postmortem-based neuroanatomical knowledge (Nieuwenhuys et al., 1988). Furthermore, previous studies based on other fibre-tracking methods described analogous trajectories for many of those tracts— the pyramidal tract Jones et al 1999, Basser

Acknowledgements

We thank Dr. Jean-Marc Vesin (Signal Processing Institute, Swiss Federal Institute of Technology, Lausanne), Eleonora Fornari, and Roberto Martuzzi (Department of Radiology, University Hospital, Lausanne) for their fruitful inputs. Thanks also to Torsten Butz for registering the MR images and to Olivier Cuisenaire for providing the visualization tool (Signal Processing Institute, Swiss Federal Institue of Technology, Lausanne).

References (41)

  • Beaulieu, C., 2002. Relating micro-structural properties of neural fibres with anisotropic diffusion, in: Proc....
  • C. Beaulieu et al.

    Determinants of anisotropic water diffusion in nerves

    Magn. Reson. Med.

    (1994)
  • M. Bozzali et al.

    White matter damage in Alzheimer’s disease assessed in vivo using diffusion tensor magnetic resonance imaging

    J. Neurol. Neurosurg. Psychiatry

    (2002)
  • S. Clarke et al.

    Forms and measures of adult and developing human corpus callosumis there sexual dimorphism?

    J. Comp. Neurol.

    (1989)
  • T. Conturo et al.

    Tracking neuronal fiber pathways in the living human brain

    Proc. Natl. Acad. Sci. USA

    (1999)
  • A. Einstein

    Investigations on the Theory of the Brownian Movement

    (1956)
  • Field, A., Alexander, A., Hasan, K., Arfanakis, K., Witwer, B., Moftakhar, R., Deshmukh, P., Haughton, V., Rowley, H.,...
  • M. Filippi et al.

    Diffusion tensor magnetic resonance imaging in multiple sclerosis

    Neurology

    (2001)
  • J. Foong et al.

    Neuropathological abnormalities of the corpus callosum in schizophreniaa diffusion tensor imaging study

    J. Neurol. Neurosurg. Psychiatry

    (2000)
  • L. Frank

    Anisotropy in high angular resolution diffusion-weighted MRI

    Magn. Reson. Med.

    (2001)
  • Cited by (276)

    View all citing articles on Scopus
    View full text