From diffusion tractography to quantitative white matter tract measures:
a reproducibility study
O. Ciccarelli,
a
G.J.M. Parker,
b
A.T. Toosy,
a
C.A.M. Wheeler-Kingshott,
a
G.J. Barker,
a
P.A. Boulby,
c
D.H. Miller,
a
and A.J. Thompson
a,
*
a
NMR Research Unit, Institute of Neurology, University College London, Queen Square, London, WC1N 3BG, UK
b
Imaging Science and Biomedical Engineering, University of Manchester, Oxford Road, Manchester M13 9PT, UK
c
MRI Unit, National Society for Epilepsy, Chalfont St. Peter, Gerrards Cross, Buckinghanshire, SL9 0RJ, UK
Received 13 March 2002; revised 12 September 2002; accepted 21 October 2002
Abstract
The aim of this study is to propose methods for assessing the reproducibility of diffusion tractography algorithms in future clinical studies
and to show their application to the tractography algorithm developed in our unit, fast marching tractography (FMT). FMT estimates
anatomical connectivity between brain regions using the information provided by diffusion tensor imaging. Three major white-matter
pathways were investigated in 11 normal subjects—anterior callosal fibers, optic radiations, and pyramidal tracts. FMT was used to generate
maps of connectivity metric, and regions of voxels with highest connectivity metric to an anatomically defined starting point were identified
for each tract under investigation. The reproducibilities of tract-“normalized” volume (NV) and fractional anisotropy (FA) measurements
were assessed over such regions. The values of tract volumes are consistent with the postmortem data. Coefficients of variation (CVs) for
FA and NV ranged from 1.7 to 7.1% and from 2.2 to 18.6%, respectively. CVs were lowest in the anterior callosal fibers (range: 1.7– 7.8%),
followed by the optic radiations (range: 1.2–18.6%) and pyramidal tracts (range: 2.6 –15.5%), suggesting that fiber organization plays a role
in determining the level of FMT reproducibility. In conclusion, these findings underline the importance of assessing the reliability of
diffusion tractography before investigating white matter pathology.
© 2003 Elsevier Science (USA). All rights reserved.
Introduction
Diffusion tensor imaging (DTI) is an MRI technique that
provides information about the random thermal motion of
water molecules in vivo (Basser et al., 1994). In human
tissues, water diffusion is not free in all directions but
hindered and restricted by the presence of barriers, includ-
ing cell cytostructure and membranes. For example, in the
white matter regions of the brain, where the neuronal pro-
jections or axons are similarly aligned, water diffusion is
generally greater in the direction parallel to axons than
perpendicular to them. This property is termed diffusion
anisotropy (Moseley et al., 1990; Chenevert et al., 1990). In
contrast, in the gray matter regions, which are characterized
by a less ordered tissue structure, diffusion tends to be less
anisotropic and more uniform in all directions. Therefore,
DTI allows in vivo identification of white matter and gray
matter regions and it has been used in these areas to study
pathological changes, including those occurring in multiple
sclerosis (Horsfield et al., 1998; Tievsky et al., 1999; Wer-
ring et al., 1999; Ciccarelli et al., 2001), tumors (Wiesh-
mann et al., 1999; Bastin et al., 1999; Inglis et al., 1999),
amyotrophic lateral sclerosis (Ellis et al., 1999), cerebral
ischemia (Chabriat et al., 1999; Sorensen et al., 1999; Jones
et al., 1999b; Helenius et al., 2002), and developmental
malformations of the cortex (Eriksson et al., 2001).
Since DTI is able to detect at the macroscopic scale of a
voxel the extent of directional bias of diffusion occurring at
the microscopic level, it can distinguish between regions
where fibers are highly aligned in the voxel from those
where fibers are less coherent. However, although DTI
provides directional information at the voxel level, it pro-
vides no explicit information about the connection between
* Corresponding author. Fax: +44-207-8136505.
E-mail address: a.thompson@ion.ucl.ac.uk (A.J. Thompson).
R
Available online at www.sciencedirect.com
NeuroImage 18 (2003) 348 –359 www.elsevier.com/locate/ynimg
1053-8119/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved.
doi:10.1016/S1053-8119(02)00042-3