First published online September 17, 2007
British Journal of Radiology (2007) 80, 782-789
© 2007 British Institute of Radiology
doi: 10.1259/bjr/99366083
White matter tractography by diffusion tensor imaging plays an important role in prognosis estimation of acute lacunar infarctions
C Lai, MD
1
S Z Zhang, MD, PhD
2
H M Liu, BSc
3
Y B Zhou, BSc
1
Y Y Zhang, MSc
1
Q W Zhang, PhD
2 and
G C Han, BSc
2
1 Department of Radiology, The Children's Hospital, 2 Department of Radiology, The Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, 3 Department of Radiology, The Longsai Hospital, Ningbo, Zhejiang, China
Correspondence: Dr Can Lai, Department of Radiology, The Children's Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, 310003 China. E-mail: laican1{at}126.com
 |
Abstract
|
|---|
The aim of this study was to evaluate the anatomical and clinical relationship between lacunar infarction and the corticospinal tract (CST) in patients with acute lacunar infarction and predict clinical outcome. We examined 28 pyramidal tract stroke patients in the acute phase or early subacute phase (<3 days) with a marked motor deficit. The anatomical location and the extent of CST involvement within the infarcts were visualized on three-dimensional colour-coded diffusion tensor tractography (DTT). With regard to the CST, all patients were divided into three clinical subgroups: Group 1 (intact type), Group 2 (partial involvement type) and Group 3 (whole involvement type). Subsequently, the severity of the motor deficit of each patient was determined according to the National Institutes of Health Stroke Scale (NIHSS) scores at the acute/early subacute phase (<3 days after onset of symptoms), early chronic phase (8–14 days) and outcome (30–60 days). NIHSS scores of Group 1 (12/28) were significantly lower than those of Group 2 (9/28) at the acute phase or early subacute phase (U = –2.816, p<0.01), and those of Group 2 were significantly lower than those of Group 3 (7/28) (U = –3.136, p<0.01). At outcome,NIHSS scores of Group 1 were significantly lower than those of Group 2 (U = –2.846, p<0.01), and scores of Group 2 were significantly lower than those of Group 3 (U = –3.130, p<0.01). At the same time, the NIHSS scores of each group gradually decreased from acute phase to outcome, Neurological improvement was statistically different among the three topographical types of infarction (H = 26.15, p<0.01; H = 11.03, p<0.01; H = 10.05, p<0.01). In conclusion, the three-dimensional colour-coded DTT allows in vivo differentiation of distinct CST stroke subtypes and may help in better establishing the prognosis for patients after CST stroke.
|
Introduction
|
|---|
Acute lacunar infarctions can be easily diagnosed on diffusion-weighted imaging (DWI). Motor weakness is a major symptom in most patients with lacunar infarction occurring in and around the internal capsule. However, on DWI, as on T2 weighted images, it is difficult to assess the precise location of the lesion with respect to the main motor sensory pathways passing through the internal capsule.
Diffusion tensor imaging (DTI) is a new imaging technique for assessing the integrity of the white matter, which normally shows anisotropic diffusion in contrast to the grey matter. The quantity and directionality of the anisotropic diffusion of the white matter can be measured and displayed as various maps (fractional anisotropy image, relative anisotropy image, colour-coded directional image). These images have been used to evaluate the anatomical relationship between the acute lacunar infarction and the internal capsule based on the existing knowledge that the CST passes through the posterior limb of the internal capsule. Three-dimensional tractography is a further development of DTI, providing exquisite anatomical definition of the white matter tract, which has yet to be applied to the study of stroke [1]. The posterior limb of the internal capsule is known to deliver the CST. In this study, it was hypothesized that the fibre tracking technique of the CST would enable a more specific localization of the lacunar infarction. Each lacunar infarction was classified into one of three detailed subtypes according to the lesion topography with regard to the CST, and was correlated topographically with the clinical symptoms.
|
Methods
|
|---|
Patients
In total, 28 consecutive patients (15 men and 13 women; mean age 63 years, range 39–83 years) admitted to our hospital because of capsular and pericapsular acute or early subacute phase (<3 days; mean 1.32 days) lacunar infarctions on DWI and isotropic imaging of DTI with a marked motor deficit were enrolled in this study. Informed consent was obtained from the patients.
DTI
DTI was conducted using a 1.5T whole-body scanner (Signa; GE Medical Systems, Milwaukee, WI) with a gradient strength of 30 mT m–1. The DTI sequence consisted of single-shot spin-echo echo-planar imaging with the following parameters: matrix 256 x 128, interpolated to 128 x 128; field of view 240 mm; slice thickness 5 mm without a gap; number of slices 24; TR 10 000 ms; TE 110 ms; number of acquisitions 2; b-value 1000 s mm–2 with 13 different directions; total scan time 4.5 min.
DTI post-processing with fibre tracking
After all of the DTI source images had been transferred to a personal computer, the diffusion tensors were calculated and fibre tracking was performed using Volume-one 1.64 and diffusion tensor visualizer 1.5 software (Department of Radiology, Tokyo University School of Medicine) based on the FACT method reported by Mori et al [2]. For fibre tracking of the CST, two regions of interest (ROIs) were manually placed by two authors on two-dimensional transverse colour-coded directional images in which the fibre bundles were coded by different colours according to their direction (red for left–right direction, green for anterior–posterior direction and blue for cranio–caudal direction); the upper ROI included the pre- and post-central gyri, the lower ROI included the pyramidal tract in the caudal portion of the pons [3, 4]. Only the fibres passing through both ROIs were displayed and designated as the CST. The threshold of the tracking termination was 0.30 for the fractional anisotropy and 0.75 for the angle between two contiguous eigen-vectors, which prohibited angles larger than 41° during tracking.
Topographical analysis with clinical symptoms
After tracking the CST, the colour-coded directional image was replaced with an isotropic diffusion image as a background image because an acute infarction was clearly shown as a bright lesion on the isotropic diffusion image. Two- and three-dimensional tractographic images were then created to localize the infarction in relation to the CST. By analysing the lesion location with regard to the CST on the tractographic images, the infarctions were visually classified into three clinical subgroups based on the extent of CST involvement by the infarct: in Group 1 lesions the CST was in close proximity to the infarction but did not pass through it; in Group 2 lesions the CST partly passed through the lesion (CST partial involvement); and in Group 3 the long lesions were centred in the pyramidal tract or involved a whole part of the CST. Each type of infarction was compared with the presence or absence of motor weakness of the three body parts (the face, the upper extremities and the lower extremities): dysarthria and/or facial weakness for the face, arm and/or hand weakness for the upper extremities and leg and/or foot weakness for the lower extremities.
Topographical analysis with clinical outcomes
Clinical neurological deficits were evaluated using the National Institutes of Health Stroke Scale (NIHSS) at the acute phase or early subacute phase (<3 days after the onset of symptoms), early chronic phase (8–14 days) and outcome (30–60 days). The interval changes of the NIHSS scores were assessed and neurological improvement was defined as a decrease of two or more points in the NIHSS score. The association between neurological improvement and the topographical type of infarction was examined by the Wilcoxon–Mann–Whitney test and the Kruskal–Wallis test.
|
Results
|
|---|
Fibre tracking of the CST was successful in all 28 patients. Group 1 infarctions (intact type), in which the CST was in close proximity to the infarction but did not pass through it, included 12 patients (Figure 1
), Group 2 (partial involvement type), in which the CST partly passed through the infarction, included 9 patients (Figure 2), and Group 3 (whole involvement type), in which long lesions were centred in the pyramidal tract or involved a whole part of the CST, included 7 patients (Figure 3). In these patients, weakness of the face, the upper extremities or the lower extremities was found, respectively. In our study, three patients with Group 1-type lesions did not have motor weakness of any part of the body and these patients were not included. However, all of these patients had sensory symptoms because the lacunar infarctions were located in the thalamus. A summary of the topographical analysis of the 28 lacunar infarctions is shown in Table 1.
View larger version (113K):
[in this window]
[in a new window]
|
Figure 1. A 52-year-old woman with an acute infarct (white arrow) in the right corona radiata 1 day after onset. (a) Axial colour-coded directional image. (b) The lesion (yellow) in the axial colour-coded directional image. (c) Three-dimensional tractography superimposed on an axial isotropic diffusion-weighted image (from an anteriosuperior viewpoint). (d) Three-dimensional tractography superimposed on an axial colour-coded directional image (from an anteriosuperior viewpoint). The corticospinal tract (orange lines) of the affected cerebral hemisphere appears to be just medial to, but not to run through, the lacunar infarct.
|
|
View larger version (108K):
[in this window]
[in a new window]
|
Figure 2. A 39-year-old man with a subacute lacunar infarct (white arrow) in the right corona radiata 3 days after onset. (a) Axial T2 weighted image. (b) Axial colour-coded directional image. (c) The lesion (yellow) in the axial colour-coded directional image. (d) Three-dimensional tractography superimposed on an axial colour-coded directional image (from an anteriosuperior viewpoint) . The corticospinal tract (orange lines) of the affected cerebral hemisphere appears to partly run through the lacunar infarct.
|
|
View larger version (95K):
[in this window]
[in a new window]
|
Figure 3. A 58-year-old woman with an acute lacunar infarct (white arrows) in the posterior limb of the right internal capsule 14 hours after onset. (a) Axial T2 weighted image. (b) Axial colour-coded directional image. (c) Three-dimensional tractography superimposed on an axial isotropic diffusion-weighted image (from an anteriosuperior viewpoint). (d) Three-dimensional tractography superimposed on an axial colour-coded directional image (from an anteriosuperior viewpoint). The corticospinal tract (orange lines) of the affected cerebral hemisphere appears to run through the lacunar infarct.
|
|
Subsequently, the severity of each patient's motor deficit was determined according to the NIHSS scores at the acute phase or early subacute phase, early chronic phase and outcome (Figure 4, Table 2). The NIHSS scores of Group 1 (12/28) were significantly lower than those of Group 2 (9/28) at the acute phase or early subacute phase (U = –2.816, p<0.01), and those of Group 2 were significantly lower than those of Group 3 (7/28) (U = –3.136, p<0.01). At the early chronic phase, the NIHSS scores of Group 1 were significantly lower than those of Group 2 (U = –3.208, p<0.01), and those of Group 2 were significantly lower than those of Group 3 (U = –3.076, p<0.01). Similarly, at outcome, the NIHSS scores of Group 1 were significantly lower than those of Group 2 (U = –2.846, p<0.01), and those of Group 2 were significantly lower than those of Group 3 (U = –3.130, p<0.01). Therefore, the NIHSS scores of Group 3 at the time of the acute or subacute phase, early chronic phase and outcome were significantly higher than those of both Groups 1 and 2.
View larger version (24K):
[in this window]
[in a new window]
|
Figure 4. Group 3(yellow line) NIHSS scores at the time of the acute or subacute phase (<3 days), early chronic phase (8–14 days) and outcome (30–60 days) are significantly higher than those of Group 1 (black line) and Group 2 (red line). NIHSS scores of Group 1 were significantly lower than those of Group 2 at the acute phase or early subacute phase, and those of Group 2 were significantly lower than those of Group 3. NIHSS scores of Group 1 were significantly lower than those of Group 2 at outcome.
|
|
The NIHSS scores of each group gradually decreased from acute phase to outcome, Neurological improvement was statistically different among the three topographical types of infarctions (H = 26.15, p<0.01; H = 11.03, p<0.01; H = 10.05, p<0.01). All members of Group 1 showed a good recovery, members of Group 2 also showed a good recovery but sometimes kept some mild motor dysfunction, and members of Group 3 always showed minor improvement and had marked deficits. Both the rADC (ADCaffected side/ADCunaffected side) value and the rFA (FAaffected side/FAunaffected side) value were induced in acute lacunar infarction (rADC 0.714, rFA 0.725).
|
Discussion
|
|---|
A small lacunar infarction occurring in and around the internal capsule is one of the most common types of ischaemic stroke. A recent article evaluated the topographical relationship between the infarction and the internal capsule in terms of the clinical prognosis using DTI [5]. In this study the colour-coded directional image created during DTI, in which the internal capsule is well recognized as a blue coloured area, was used to easily localize the lacunar infarction with respect to the internal capsule. However, it was not possible to separate the CST from the internal capsule on the colour-coded directional image alone, even though the CST is well known to lie in the posterior third quarter of the posterior limb of the internal capsule. The fibre tracking technique is the only method that enables a precise topographical evaluation of the lacunar infarction and the CST [6].
The results of the fibre tracking concerning the presence or absence of fibre involvement by the lesion are difficult to validate. A few clinical studies of lacunar infarctions and brain tumours have validated the results of fibre tracking based on patient symptoms [7, 8]. The relationship between the position of the infarct and clinical symptoms was highly correlated. The degree of involvement, however, cannot be quantified from the current data. A method that allows quantification or accurate grading of tract involvement would be an important step and warrants further investigation.
In our study, the anatomical location and the extent of CST involvement within the infarcts were visualized on DTT. With regard to the CST, all patients were divided into three clinical subgroups (e.g. whole involvement), and this classification was compared with the severity of the patients' motor deficits according to NIHSS scores at the acute phase or early subacute phase, early chronic phase and outcome. NIHSS scores of the whole involvement-type subgroup at the time of the acute or subacute phase, early chronic phase and outcome were significantly higher than those of the intact-type and partial involvement-type subgroups. Scores of the partial involvement-type subgroup were significantly higher than those of the intact-type subgroup. At the same time, the NIHSS scores of each subgroup gradually decreased from acute phase to outcome. Neurological improvement was statistically different among the three topographical types of infarctions. Therefore, the good correlation between the location of the lesion and the patient's symptoms in this study supports the topographical accuracy of the fibre tracking of the CST. These results also increase the feasibility of using this technique in such a way as to expand its clinical utility.
There may be a concern as to the reproducibility of the fibre tracking. The CST is composed of over one million fibres arising from the pre-central gyrus (Brodmann area four), the pre-motor cortex (area six), the post-central gyrus (area one, two and three) and the adjacent parietal cortex (area five) [9]. These fibres converge into the corona radiata and then form a longitudinally oriented compact white matter bundle with a high anisotropy. Therefore, the shape of the CST from the level of the internal capsule to the level of the brainstem is rather simple compared with the other white matter tracts, and its location can be easily estimated. It is believed that the fibre tracking method of the CST, as used in this study, is easy to perform and that the results are highly reproducible [10].
Fibre crossing reduces the anisotropy of the voxel because of the averaging of the fibres with different directions, which hampers the fibre tracking. The laterally oriented fibres of the CST in the region above the level of the corona radiata are crossed by the fibres of the superior longitudinal fasciculus. Therefore, the fibres of the CST arising from the inferior frontoparietal cortices (i.e. the face and tongue area) are difficult to track when the ROI for fibre tracking is placed at or laterally to the superior longitudinal fasciculus toward the cortex. However, we placed two ROIs at the brainstem inferiorly and at the internal capsule superiorly to minimize the fibre crossing effect in this study.
Recently, an article reported a significant correlation between patient outcome and the degree of CST involvement by the infarction, although there was a slight difference in the methodology used for depicting the CST and analysing the outcome compared with this study [11]. Further studies using larger groups of patients are needed to examine the clinical significance of the topographical classification in terms of the patient's outcome.
|
Conclusion
|
|---|
This study investigated the ability of the fibre tracking technique to localize lacunar infarctions, particularly with respect to the body parts affected, by correlating the lesion location with the clinical symptoms topographically. The results suggest that this technique is able to specifically localize capsular, pericapsular and corona radiata infarctions with regard to the body parts affected. In addition, the results also strongly support the topographical accuracy of the fibre tracking of the CST.
Received for publication December 2, 2006.
Revision received December 6, 2006.
Accepted for publication January 2, 2007.
|
References
|
|---|
- Osamu A, Yoshitaka M, Shigeki A, Hidenori Y, Haruyasu Y, Kiyoto K, et al. Topography of the human corpus callosum using diffusion tensor tractography. J Comput Assist Tomogr 2004;28:533–9.[CrossRef][Medline]
- Mori S, Crain BJ, Chacko VP, van Zijl PC. Three-dimensional tracking of axonal projections in the brain by magnetic resonance imaging. Anu Neurol 1999;45:265–9.[CrossRef]
- Setsu W, Hang YJ, Lindia M, van Zijl PC, Mori S. Fiber tract-based atlas of human white matter anatomy. Radiology 2004;230:77–87.[Abstract/Free Full Text]
- Iwasaki S, Nakagawa H, Fukusmi A, Kichikawa K, Kitamura K, Otsuji H, et al. Identification of pre- and postcentral gyri on CT and MR images on the basis of the medullary pattern of cerebral white matter. Radiology 1991;179:207–13.[Abstract/Free Full Text]
- Lie C, Hirsch JG, Rossmanith C, Hennerici MG, Gass A. Clinicotopographical correlation of corticospinal tract stroke: a color-coded diffusion tensor imaging study. Stroke 2004;351:92–3.
- Jong SL, Moon KH, Sung HK, Kwon OK, Jae HK. Fiber tracking by diffusion tensor imaging in corticospinal tract stroke: topographical correlation with clinical symptoms. NeuroImage 2005;26:771–6.[CrossRef][Medline]
- Yamada K, Mori S, Nakamura H, Ito H, Kizu O, Shiga K, et al. Fiber-tracking method reveals sensorimotor pathway involvement in stroke patients. Stroke 2003;34:e159–e162.[Abstract/Free Full Text]
- Yamada K, Ito H, Nakamura H, Osamu K, Wataru A, Takao K, et al. Stroke patients' evolving symptoms assessed by tractography. J Ragn Reson Imaging 2004;20:923–9.[CrossRef]
- Carpenter MB, Sutin J. Human neuroanatomy, 8th edn. Baltimore, MD: Williams and Wilkins, 1983. 282–3.
- Kunimatsu A, Aoki S, Masutani, Y, Abe O, Mori H, Ohtomo K, et al. Three-dimensional white matter tractography by diffusion tensor imaging in ischaemic stroke involving the corticospinal tract. Neuroradiology 2003;45:532–5.[CrossRef][Medline]
- Konishi J, Yamada K, Kizu O, Ito H, Sugimura K, Yoshikawa K, et al. MR tractography for the evaluation of functional recovery from lenticulostriate infarcts. Neurology 2005;64:108–13.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
The STEPS Participants
Stem Cell Therapies as an Emerging Paradigm in Stroke (STEPS): Bridging Basic and Clinical Science for Cellular and Neurogenic Factor Therapy in Treating Stroke
Stroke,
February 1, 2009;
40(2):
510 - 515.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
BJR review of the year -- 2007
Br. J. Radiol.,
April 1, 2008;
81(964):
265 - 269.
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
