Professional Documents
Culture Documents
1 s2.0 S0720048X09002174 Main
1 s2.0 S0720048X09002174 Main
a r t i c l e
i n f o
Article history:
Received 8 January 2009
Received in revised form 6 April 2009
Accepted 6 April 2009
Keywords:
Tensor diffusion trace-weighted imaging
Acute cerebral infarction
a b s t r a c t
Background and purpose: Although isotropic diffusion-weighted imaging (isoDWI) is very sensitive to the
detection of acute ischemic stroke, it may occasionally show diffusion negative result in hyper-acute
stroke. We hypothesize that high diffusion contrast diffusion trace-weighted image with enhanced T2
may improve stroke lesion conspicuity.
Methods: Five hyper acute stroke patients (M:F = 0:5, average age = 61.8 20.5 y/o) and 16 acute
stroke patients (M:F = 11:5, average age = 67.7 12 y/o) were examined six-direction tensor DWIs at
b = 707 s/mm2 . Three different diffusion-weighted images, including isotropic (isoDWI), diffusion traceweighted image (trDWI) and T2-enhanced diffusion trace-weighted image (T2E trDWI), were generated.
Normalized lesion-to-normal ratio (nLNR) and contrast-to-noise ratio (CNR) of three diffusion images
were calculated from each patient and statistically compared.
Results: The trDWI shows better nLNR than isoDWI on both hyper-acute and acute stroke lesions, whereas
no signicant improvement in CNR. Nevertheless, the T2E trDWI has statistically superior CNR and nLNR
than those of isoDWI and trDWI in both hyper-acute and acute stroke.
Conclusions: We concluded that tensor diffusion trace-weighted image with T2 enhancement is more
sensitive to stroke lesion detection, and can provide higher lesion conspicuity than the conventional
isotropic DWI for early stroke lesion delineation without the need of high-b-value technique.
2009 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
Diffusion-weighted image (DWI) has been proven the most sensitive diagnostic as well as outcome prediction tool for the acute
ischemic stroke in the rst few hours without the need of contrast
agent [110]. Isotropic (traced) DWI is superior to anisotropic DWI
as post-processing isotropic DWIs help to avoid misinterpreting the
stroke lesion by eliminating the diffusion anisotropy from the normal white matter tracts [11]. Notwithstanding, hyper-acute stroke
could exhibit as diffusion negative stroke in isotropic DWI [1216].
Therefore, there is room for improvement in the diffusion imaging method in order to increase the sensitivity of DWI in stroke
This study was supported in part by Chi-Mei National Defense Medical Center
grant CNNDMC9508 and Tri-Service General Hospital grant TSGH-C93-19.
Corresponding author at: Department of Radiology, Tri-Service General Hospital,
325, Cheng-Kung Road, NeiHu, Taipei, Taiwan, ROC.
E-mail address: sandy0928@seed.net.tw (C.-Y. Chen).
0720-048X/$ see front matter 2009 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.ejrad.2009.04.023
e90
trDWI
1/2 12
= S0
= S0 exp
D +D +D
x
y
z
where, DWIi,j is the diffusion image with diffusion sensitizing gradients applied in (i,j) direction, and S0 is the image acquired without
diffusion sensitizing gradient.
Apart from tensor isotropic DWI, diffusion trace-weighted image
(trDWI) [11] was also calculated according to the following equation,
(1)
1/2 4
= S0
= S0 exp ((Dx + Dy + Dz ) b)
(2)
4
3/2
= S0 exp ((Dx + Dy + Dz ) b)
(3)
In this equation, an order of 3/2 of S0 is used. Since S0 is a T2weighted image, with the order larger than identity, the DWI will
appear higher T2 contrast. In general, the order of S0 can be any
value larger than identity, but it will cause dramatically heavy T2weighted image if larger order of S0 was chosen. For convenience
in this study, we use order of 3/2 for S0 which is directly derived by
taking quarter root of six multiplied DWIs.
2.3. Data analysis
To compare three different DWIs, normalized lesion-to-normal
ratio (nLNR) and contrast-to-noise ratio (CNR) were calculated by
Table 1
Demographics and clinical characteristics of 21 patients with acute ischemic stroke.
Patient no.
Gender
Age (years)
MRA ngings
Timea
Symptom
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
M
M
M
M
M
F
F
M
F
M
F
M
M
F
F
M
F
M
F
F
F
69
75
61
45
81
26
73
52
47
72
71
80
78
79
73
74
55
77
68
63
74
Negative
R ICA and MCA occlusion
R ICA and MCA occlusion
L M1 Occlusion
R PCA occlusion
L M2 narrowing
R M1 occlusion
L MCA occlusion
L M1/2 and P1/2 narrowing
R MCA and PCA narrowing
L ACA occlusion
L P1 and R M2 narrowing
R MCA occlusion
R ICA and MCA occlusion
Negative
R M2 and P1/2 narrowing
R M2 Occlusion
L ICA and MCA occlusion
R MCA occlusion
L MCA occlusion
L M1 narrowing
42
42
47
39
48
5
4
48
24
30
72
72
66
11
5
72
40
72
24
3
6
Note: MCA, middle cerebral artery; ICA, internal carotid artery; ACA, anterior cerebral artery; PCA, posterior cerebral artery; M1, M1 segment of MCA; M2, M2 segment of
MCA; P1, P1 segment of posterior cerebral artery; R, right; L, left.
a
Symptom to MRI time (h); MRA, magnetic resonance angiography.
e91
3. Results
3.1. Hyper-acute stroke
Fig. 2. Plot of contrast-to-noise ratio (CNR) calculated on three different diffusionweighted images in ve hyper-acute stroke patients. The asterisk indicates P < 0.05.
CNR =
Slesion Snormal
Slesion + Snormal
Slesion Snormal
Noise
(4)
(5)
Fig. 3. Three different diffusion-weighted images and one FLAIR image of a 63 years old female stroke patient with symptom onset within 3 h. The images from the left to
right are trace ADC (a), conventional isoDWI (b), trDWI (c), T2E trDWI (d) and FLAIR (e). White arrows point out the margin of stroke lesion which cannot be delineated by
isoDWI (b), but clearly dened by both trDWI (c) and T2E trDWI (d).
e92
Fig. 4. Four slices of trace ADC map and three different diffusion-weighted images of a 74 years old stroke patient with symptom onset within 6 h. In this case, the core
stroke lesion, which is located in slice (d), can be visualized by trace ADC map (1st column), isoDWI (2nd column), trDWI (3rd column) and T2E trDWI (4th column). However,
the diffusion negative signal, which occurs at the suburbs of central stroke lesion in isoDWIs (rows (a)(c) of 2nd column), could be visualized in trDWIs (rows (a)(c) of 3rd
column) and T2E trDWIs (rows (a)(c) of 4th column) indicated by white arrows.
stroke patient groups. However, to achieve signicant, not quantum, imaging improvement for DWI, a very high-b-value up to
20003500 s/mm2 is usually required [1719]. This is usually at
the expense of signal to noise ratio and is not always applicable
in routine MR scanner due to the limited gradient strength.
Fig. 6. Plot of contrast-to-noise ratio (CNR) calculated on three different diffusionweighted images in 16 acute stroke patients. The asterisk indicates P < 0.05.
e93
Fig. 7. Trace ADC maps and three diffusion-weighted images of three slices from a 77-year-old stroke patient with symptom onset within 72 h. In this case, the scattered
stroke lesions of the rst slice were visualized in trace ADC map (1st column), isoDWI (2nd column), trDWI (3rd column) and T2E trDWI (4th column). However, several small
lesions, showing diffusion negative signal in isoDWI (rows (b) and (c)), were observed in trDWI and T2E trDWI as indicated by white arrows.
contrast from the trDWI and T2E trDWI, especially in cerebral ventricles showing the intermediate signal intensity, which caused the
adjacent tiny stroke lesions inconspicuous in isoDWI, as shown in
Fig. 7(rows (b) and (c)). By taking advantage of increasing diffusion
exponential term, the trDWI and T2E trDWI were both capable of
contrasting tiny stroke lesions in the vicinity of cerebral ventricles.
However, it should be noticed that although our study showed
that T2E trDWI helps improve CNR by utilizing heavily T2 and diffusion weightings in diagnosis of cerebral infarct, the trace ADC
maps, i.e. sum of three orthogonal diffusivities, of isoDWI, trDWI and
T2E trDWI were numerically identical. Therefore, trace ADC maps
of isoDWI, trDWI and T2E trDWI were not compared in this study.
5. Conclusions
In this study, we demonstrated that tensor trDWI helps suppress background tissue and therefore enhances stroke lesion on
both acute and hyper-acute stroke patients. With T2 enhancement,
T2E trDWI helps improve the conspicuity of acute stroke lesions
without the need of high gradient strength (or high-b-value) in
routine MR scanners.
References
[1] Gonzalez RG, Schaefer PW, Buonanno FS, et al. Diffusion-weighted MR imaging:
diagnostic accuracy in patients imaged within 6 h of stroke symptom onset.
Radiology 1999;210:15562.
[2] Singer MB, Chong J, Lu DF, Schonewille WJ, Tuhrim S, Atlas SW. Diffusionweighted MRI in acute subcortical infarction. Stroke 1998;29:1336.
[3] Warach S, Chien D, Li W, Ronthal M, Edelman RR. Fast magneticresonance diffusion-weighted imaging of acute human stroke. Neurology
1992;42:171723.
e94
[4] Hand PJ, Wardlaw JM, Rivers CS, et al. MR diffusion-weighted imaging and
outcome prediction after ischemic stroke. Neurology 2006;66:115963.
[5] Kuhl CK, Textor J, Gieseke I, et al. Acute and subacute ischemic stroke at
high-eld-strength (30-T) diffusion-weighted MR imaging: intraindividual
comparative study. Radiology 2005;234:50916.
[6] Singhal AB, Topcuoglu MA, Koroshetz WJ. Diffusion MRI in three types of anoxic
encephalopathy. Journal of the Neurological Sciences 2002;196:3740.
[7] Arbelaez A, Castillo M, Mukherji SK. Diffusion-weighted MR imaging of global
cerebral anoxia. American Journal of Neuroradiology 1999;20:9991007.
[8] Pereira RS, Harris AD, Sevick RJ, Frayne R. Effect of b value on contrast during
diffusion-weighted magnetic resonance imaging assessment of acute ischemic
stroke. Journal of Magnetic Resonance Imaging 2002;15:5916.
[9] van Everdingen KJ, van der Grond J, Kappelle LJ, Ramos LMP. Mali WPTM.
Diffusion-weighted magnetic resonance imaging in acute stroke. Stroke
1998;29:178390.
[10] Warach S, Gaa J, Siewert B, Wielopolski P, Edelman RR. Acute human stroke
studied by whole-brain echo-planar diffusion-weighted magnetic-resonanceimaging. Annals of Neurology 1995;37:23141.
[11] Chong JN, Lu DF, Aragao F, et al. Diffusion-weighted MR of acute cerebral
infarction: comparison of data processing methods. American Journal of Neuroradiology 1998;19:17339.
[12] Oppenheim C, Stanescu R, Dormont D, et al. False-negative diffusion-weighted
MR ndings in acute ischemic stroke. American Journal of Neuroradiology
2000;21:143440.
[13] Ay H, Buonanno FS, Sorenson G, et al. Negative diffusion-weighted MR imaging
(DWI) with stroke-like focal decits. Stroke 1998;29:3071307.
[14] Crozier S, Oppenheim C, Stanescu R, et al. Early negative diffusion
MRI (DWI) in ischemic vertebro-basilar stroke. Neurology 2000;54:A94
194.