Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print October 19, 2006, 10.1148/radiol.2413051182
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2413051182v1
241/3/873    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pouwels, P. J. W.
Right arrow Articles by Barkhof, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Pouwels, P. J. W.
Right arrow Articles by Barkhof, F.
(Radiology 2006;241:873-879.)
© RSNA, 2006

Technical Developments

Human Gray Matter: Feasibility of Single-Slab 3D Double Inversion-Recovery High-Spatial-Resolution MR Imaging1

Petra J. W. Pouwels, PhD, Joost P. A. Kuijer, PhD, John P. Mugler, III, PhD, Charles R. G. Guttmann, MD and Frederik Barkhof, MD, PhD

1 From the MS Research Center, Department of Physics and Medical Technology (P.J.W.P., J.P.A.K.) and Department of Radiology (F.B.), VU University Medical Center, De Boelelaan 1117, 1081 HV Amsterdam, the Netherlands; Department of Radiology, University of Virginia School of Medicine, Charlottesville, Va (J.P.M.); and Center for Neurological Imaging, Departments of Radiology and Neurology, Brigham and Women's Hospital, Harvard Medical School, Boston, Mass (C.R.G.G.). Received July 14, 2005; revision requested September 21; revision received November 29; accepted January 10, 2006; final version accepted February 13. P.J.W.P., J.P.A.K., and F.B. supported by the Dutch MS Research Foundation (Stichting MS Research), Voorschoten, the Netherlands. J.P.M. and C.R.G.G. supported by the National Institutes of Health (R01 NS035142). C.R.G.G. supported by the National Multiple Sclerosis Society (RG3574-A-1). Address correspondence to P.J.W.P. (e-mail: pjw.pouwels{at}vumc.nl).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
The purpose of this study was to develop and prospectively evaluate the feasibility of a single-slab three-dimensional (3D) double inversion-recovery, or DIR, sequence for magnetic resonance imaging at 1.5 T. The study was approved by the local ethics committee, and informed consent was obtained from six healthy control subjects (one woman, five men; age range, 26–47 years) and two patients with multiple sclerosis (one woman, aged 39; one man, aged 56). Gray matter (GM)–only images were obtained by selectively suppressing cerebrospinal fluid (CSF) and white matter (WM) signals. Whole-brain high-spatial-resolution 3D images (1.2 x 1.2 x 1.3 mm) were acquired within 10 minutes. Cortical and deep GM structures were clearly delineated from WM and CSF, and there were regional differences in GM signal intensity. No flow artifacts from blood or CSF were observed. These GM images with high spatial resolution are suitable to identify cortical pathologic conditions and can potentially be used for segmentation purposes to determine cortical thickness or volume.

© RSNA, 2006


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Selectively imaging gray matter (GM) structures in the human brain is important in the study of many neurologic diseases. The determination of total cortical volume can be used to assess an overall degree of cortical atrophy (1,2), and the determination of local cortical thickness can be used to detect regional abnormalities, such as those that occur with epilepsy (3,4). When distinct cortical structures are prone to degeneration (eg, the hippocampus and the entorhinal cortex in patients with Alzheimer disease), it is important to reliably assess volume or thickness of these small structures, especially for longitudinal studies such as clinical trials (2,5). In patients with multiple sclerosis (MS), the identification of cortical lesions has been shown to improve when selective GM imaging techniques are used (6).

Two-dimensional double inversion-recovery (DIR) magnetic resonance (MR) imaging techniques have been proposed to selectively depict GM by suppressing the signals from both white matter (WM) and cerebrospinal fluid (CSF) (7–9). Because of the thin and folded structure of the cortex, high-spatial-resolution three-dimensional (3D) MR imaging is preferable to two-dimensional MR imaging, which typically employs relatively thick sections. A few studies have reported the use of multislab 3D DIR MR pulse sequences with a high in-plane spatial resolution (1.0 x 1.0 mm to 1.2 x 1.2 mm) and improved through-plane resolution when compared with those of two-dimensional MR imaging (1.75–2.0 mm) (6,10,11). A disadvantage of these multislab 3D sequences is the requirement for oversampling in the slab direction to compensate for imperfect slab profiles and to prevent wraparound artifacts; nonetheless, appropriate interleaving of the slabs allows an efficient use of the acquisition duration. Furthermore, slab-to-slab radiofrequency-field homogeneity differences cause a "venetian blind" artifact, although this can be minimized with optimal parameter adjustments (10). And finally, flow artifacts from blood and CSF cannot completely be avoided with multislab 3D sequences (6,10,12).

Single-slab 3D MR sequences (13,14) have a long echo train and variable flip angles for refocusing radiofrequency pulses, which yield high-quality images with whole-brain coverage. The sequences have been implemented with several contrasts: T1 weighted, T2 weighted, and fluid-attenuated inversion recovery (FLAIR) (15,16). The purpose of this study was to develop and prospectively evaluate the feasibility of a single-slab 3D DIR sequence for MR imaging at 1.5 T.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Subjects
This study was performed with six healthy control subjects (one woman, five men; age range, 26–47 years) and two patients with MS (one woman, aged 39 years; one man, aged 56 years). Patients were randomly selected from a clinical database of patients with MS. Health status of the six control subjects was determined by evaluating their medical history. The study protocol was approved by the local ethics committee, and subjects gave informed consent prior to undergoing MR examination.

MR Imaging and Evaluation
Examinations were performed with a 1.5-T whole-body MR imager (Sonata; Siemens, Erlangen, Germany) with the standard circularly polarized head coil. All inversion, excitation, and refocusing radiofrequency pulses in the 3D DIR sequence were non–spatially selective. The sequence had the following layout: The time between the two adiabatic inversion pulses was the inversion time (TI) TIlong, and the time between the second inversion pulse and the excitation pulse was TIshort. The readout period consisted of a long spin-echo train with 191 echoes. Refocusing radiofrequency pulses with variable flip angles were employed to achieve a signal evolution for GM, which started with an exponential decay, was followed by a constant level, and ended with an exponential decay as described previously (14). This design yielded a narrow point spread function for GM and did not introduce blurring despite the long echo train (14). The bandwidth was 490 Hz/pixel, the readout period was 700 msec, the echo spacing was 3.7 msec, and the effective echo time (ie, the time of sampling the center of k-space with respect to the excitation pulse) was 355 msec. Because the prescribed signal evolution contained a substantial contribution from magnetization stored along the z-axis, the image appearance is much different from that which would be obtained with an equivalent echo time in a conventional spin-echo sequence. With a T2-weighted pulse sequence configuration, the chosen readout period and effective echo time yielded contrast similar to that provided by a conventional spin-echo pulse sequence with an echo time of approximately 100 msec (14). The excitation radiofrequency pulse was preceded by a fat saturation pulse and an inferiorly placed saturation slab to minimize flow artifacts. Reconstruction was performed in magnitude mode.

The repetition time was fixed at 6500 msec, which was in the optimal range for the signal-to-noise ratio of GM (SNRGM) if imaging time is taken into account (10). TIs were empirically optimized (P.J.W.P.)—on the basis of initial calculations similar to those described by Boulby et al (10)—at examinations in three control subjects (randomly selected from the six control subjects). TIlong was 2300 –2550 msec in intervals of 50–100 msec, and TIshort was 335–370 msec in intervals of 20–10 msec. Optimal suppression of both CSF and WM signals while retaining GM signals, as determined visually and with measurements of signal intensity, was obtained with a TIlong of 2350 msec and a TIshort of 350 msec, and these TIs were used at subsequent 3D DIR examinations.

Whole-brain coverage was obtained with a sagittal 3D slab of 156 mm, which consisted of 120 partitions of 1.3 mm. By using a 190 x 256 matrix with readout in the head-to-feet direction (to prevent infolding) and a 230 x 310-mm field of view, the in-plane resolution was 1.21 x 1.21 mm. No interpolation was used either in plane or in the slab direction. By employing 75% partial Fourier in the slab direction, the acquisition time was 9 minutes 47 seconds.

Three-dimensional DIR MR acquisitions were obtained in the other three control subjects. These images were assessed qualitatively by visual inspection (F.B., 15 years of experience in neuroradiology, in consensus with P.J.W.P., 10 years of experience in MR physics). The overall degree of WM and CSF signal suppression was judged by comparing the signal intensity of WM with the signal intensity of CSF and comparing signal intensities of WM and CSF with that of air in the background. The signal intensity of GM was assessed for regional variations. Images were examined for the presence of high signal intensity in arteries, the choroid plexus, and ventricular borders, which is indicative of flow artifacts from blood or CSF. Reconstructed images in other orientations were judged for the presence of section-to-section signal intensity variations along the slab-select direction.

Signal-to-noise ratios of GM (SNRGM), WM (SNRWM), and CSF (SNRCSF) were determined (P.J.W.P.) by dividing the signal intensity in small regions of interest by the standard deviation (SD) of the noise (SDnoise). Regions of interest in selected GM structures (hippocampus, amygdala, cingulate gyrus, and motor cortex) had areas of 18–36 mm2. Regions of interest in WM (corpus callosum and frontal WM) and CSF (lateral ventricles) had areas of 18–44 mm2. SDnoise was determined by measuring the SD of the pixel intensities in a background region of air (with an area of 120 mm2) containing no visible signal intensity; this value was then corrected to obtain the true SD on the basis of the Rayleigh distribution that governs the noise in regions with no signal intensity (17,18). Contrast-to-noise ratios (CNRs) of GM versus WM (CNRGM-WM) and of GM versus CSF (CNRGM-CSF) were determined by dividing the difference in signal intensity between GM and WM or between GM and CSF by SDnoise. Values are given as mean ± SD.

For clinical illustration, 3D DIR MR acquisitions were also obtained in two patients with MS. These images were judged for the presence of lesions in GM and WM. The degree of WM and CSF signal suppression was assessed qualitatively by visual inspection (F.B., P.J.W.P.) of the images as described. In addition, SNR and CNR values were calculated as defined above.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Control Subjects
Uniform suppression of WM and CSF was observed on single-slab 3D DIR images in the original sagittal orientation in the three control subjects (Fig 1). The degree of WM suppression was homogeneous throughout the brain; no difference in signal intensity between the corpus callosum and periventricular WM was seen. The only exception was the WM extending from the central sulcus to the thalamus, for which the suppression was slightly less efficient (Fig 1b). On sagittal views through the midline (Fig 1a), the pons was well delineated because it had a lower signal intensity than did the surrounding structures of the brainstem. All cortices and deep GM structures were well delineated, and the septum pellucidum and choroid plexus also had high signal intensity. There was regional variation in GM signal intensity; the highest signal intensity was in the limbic lobe (cingulate gyrus, hippocampus, and amygdala) and the lowest signal intensity was in the cortex around the central sulcus.


Figure 1
View larger version (160K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a: (a–d) Sagittal single-slab 3D DIR MR images of a control subject (repetition time/echo time msec, 6500/355; TIlong/TIshort, 2350/350; field of view, 230 x 310 mm; matrix, 190 x 256; section thickness, 1.3 mm). Both CSF and WM signals are well suppressed, resulting in clear delineation of GM structures. Note differences in GM signal intensity, which is highest in cingulate cortex (arrow in a), amygdala, and hippocampus (arrow in b) and lowest in cortex along central sulcus (arrowhead in b and c).

 

Figure 1
View larger version (155K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b: (a–d) Sagittal single-slab 3D DIR MR images of a control subject (repetition time/echo time msec, 6500/355; TIlong/TIshort, 2350/350; field of view, 230 x 310 mm; matrix, 190 x 256; section thickness, 1.3 mm). Both CSF and WM signals are well suppressed, resulting in clear delineation of GM structures. Note differences in GM signal intensity, which is highest in cingulate cortex (arrow in a), amygdala, and hippocampus (arrow in b) and lowest in cortex along central sulcus (arrowhead in b and c).

 

Figure 1
View larger version (145K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1c: (a–d) Sagittal single-slab 3D DIR MR images of a control subject (repetition time/echo time msec, 6500/355; TIlong/TIshort, 2350/350; field of view, 230 x 310 mm; matrix, 190 x 256; section thickness, 1.3 mm). Both CSF and WM signals are well suppressed, resulting in clear delineation of GM structures. Note differences in GM signal intensity, which is highest in cingulate cortex (arrow in a), amygdala, and hippocampus (arrow in b) and lowest in cortex along central sulcus (arrowhead in b and c).

 

Figure 1
View larger version (131K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1d: (a–d) Sagittal single-slab 3D DIR MR images of a control subject (repetition time/echo time msec, 6500/355; TIlong/TIshort, 2350/350; field of view, 230 x 310 mm; matrix, 190 x 256; section thickness, 1.3 mm). Both CSF and WM signals are well suppressed, resulting in clear delineation of GM structures. Note differences in GM signal intensity, which is highest in cingulate cortex (arrow in a), amygdala, and hippocampus (arrow in b) and lowest in cortex along central sulcus (arrowhead in b and c).

 
Artifacts
Artifacts from blood flow or CSF pulsation were not visually apparent on the images. For example, the absence of CSF flow artifacts can be well appreciated in Figure 1a because of the very low signal intensity in the aqueduct and in the fourth ventricle. There were no pulsation artifacts caused by blood flow, which would otherwise be expected, especially in the sagittal sinus. The absence of flow artifacts was predominantly a result of the nonselective excitation of the single-slab sequence. Acquisition of the data without use of the inferiorly placed saturation slab did not result in a detectable increase of flow artifacts (data not shown).

The use of single-slab excitation resulted in homogeneous signal intensity along the slab-select direction without section-to-section differences (Fig 2). The reconstructed images have the same spatial resolution as the original sagittal images, and there is no image degradation due to interpolation effects of nonisotropic pixels.


Figure 2
View larger version (125K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: (a, b) Coronal and (c, d) transverse reconstructed MR images of single-slab 3D DIR MR acquisition in a control subject (6500/355; TIlong/TIshort, 2350/350). Because of near-isotropic resolution, image quality is similar in all orientations. Note clear regional variation of GM signal intensity, such as difference between cingulate cortex (arrow) and motor cortex (arrowheads) in c.

 

Figure 2
View larger version (124K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: (a, b) Coronal and (c, d) transverse reconstructed MR images of single-slab 3D DIR MR acquisition in a control subject (6500/355; TIlong/TIshort, 2350/350). Because of near-isotropic resolution, image quality is similar in all orientations. Note clear regional variation of GM signal intensity, such as difference between cingulate cortex (arrow) and motor cortex (arrowheads) in c.

 

Figure 2
View larger version (76K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c: (a, b) Coronal and (c, d) transverse reconstructed MR images of single-slab 3D DIR MR acquisition in a control subject (6500/355; TIlong/TIshort, 2350/350). Because of near-isotropic resolution, image quality is similar in all orientations. Note clear regional variation of GM signal intensity, such as difference between cingulate cortex (arrow) and motor cortex (arrowheads) in c.

 

Figure 2
View larger version (135K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2d: (a, b) Coronal and (c, d) transverse reconstructed MR images of single-slab 3D DIR MR acquisition in a control subject (6500/355; TIlong/TIshort, 2350/350). Because of near-isotropic resolution, image quality is similar in all orientations. Note clear regional variation of GM signal intensity, such as difference between cingulate cortex (arrow) and motor cortex (arrowheads) in c.

 
SNRs and CNRs
For regions of interest placed in the hippocampus, amygdala, and cingulate gyrus, the mean SNRGM was 18.3 ± 1.2. The lower signal intensity in the cortex along the central sulcus led to a lower SNRGM of 9.8 ± 1.0. The suppression of CSF and WM signals was equally efficient and very similar in the three control subjects, which yielded an SNRCSF of 2.8 ± 0.6 and an SNRWM of 2.7 ± 0.6. For CNRs, the GM with the highest signal intensity was used, which resulted in a CNRGM-CSF of 15.8 ± 0.9 and a CNRGM-WM of 15.6 ± 1.6.

Patients with MS
The single-slab 3D DIR MR images of the two patients with MS (Fig 3) illustrate the high signal intensity of lesions that was obtained with this sequence. In particular, cortical lesions could be detected and differentiated from WM lesions because of the good distinction between GM and WM. WM lesions, such as those in periventricular WM, in the corpus callosum, and in the brainstem, also could be identified. Whereas the degree of CSF suppression on images of the patients with MS was similar to that on the images of the control subjects, the suppression of WM on images of patients with MS was not as efficient as that on images of control subjects. The signal intensity of WM on the 3D DIR images of patients with MS was slightly higher and more variable than that of the control subjects. These qualitative observations were confirmed with the analysis of SNR and CNR on images of these patients: SNRGM = 18.5 ± 1.6, SNRCSF = 3.1 ± 0.3, and SNRWM = 5.2 ± 1.7, whereas CNRGM-CSF = 14.4 ± 1.0, and CNRGM-WM = 13.3 ± 2.3.


Figure 3
View larger version (140K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a: Single-slab 3D DIR MR images of two patients with MS for clinical illustration (6500/355; TIlong/TIshort, 2350/350). (a, b) Sagittal images of a 39-year-old patient show hyperintense lesions in cortex (some indicated with arrowheads), corpus callosum (upper arrow), and brainstem (lower arrow). (c) Coronal and (d) transverse reconstructed images of a 56-year-old patient show predominantly WM lesions (some indicated with arrows). Slightly higher signal intensity of WM with respect to CSF, especially evident in coronal image, is probably a result of increased T1 relaxation times of normal-appearing WM, which led to incomplete WM suppression.

 

Figure 3
View larger version (65K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b: Single-slab 3D DIR MR images of two patients with MS for clinical illustration (6500/355; TIlong/TIshort, 2350/350). (a, b) Sagittal images of a 39-year-old patient show hyperintense lesions in cortex (some indicated with arrowheads), corpus callosum (upper arrow), and brainstem (lower arrow). (c) Coronal and (d) transverse reconstructed images of a 56-year-old patient show predominantly WM lesions (some indicated with arrows). Slightly higher signal intensity of WM with respect to CSF, especially evident in coronal image, is probably a result of increased T1 relaxation times of normal-appearing WM, which led to incomplete WM suppression.

 

Figure 3
View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3c: Single-slab 3D DIR MR images of two patients with MS for clinical illustration (6500/355; TIlong/TIshort, 2350/350). (a, b) Sagittal images of a 39-year-old patient show hyperintense lesions in cortex (some indicated with arrowheads), corpus callosum (upper arrow), and brainstem (lower arrow). (c) Coronal and (d) transverse reconstructed images of a 56-year-old patient show predominantly WM lesions (some indicated with arrows). Slightly higher signal intensity of WM with respect to CSF, especially evident in coronal image, is probably a result of increased T1 relaxation times of normal-appearing WM, which led to incomplete WM suppression.

 

Figure 3
View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3d: Single-slab 3D DIR MR images of two patients with MS for clinical illustration (6500/355; TIlong/TIshort, 2350/350). (a, b) Sagittal images of a 39-year-old patient show hyperintense lesions in cortex (some indicated with arrowheads), corpus callosum (upper arrow), and brainstem (lower arrow). (c) Coronal and (d) transverse reconstructed images of a 56-year-old patient show predominantly WM lesions (some indicated with arrows). Slightly higher signal intensity of WM with respect to CSF, especially evident in coronal image, is probably a result of increased T1 relaxation times of normal-appearing WM, which led to incomplete WM suppression.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 References
 
The single-slab 3D MR turbo spin-echo pulse sequence, which has previously been described with T1-weighted, T2-weighted, and FLAIR contrasts, has now been described with a DIR contrast. This single-slab 3D DIR method produces images that selectively depict GM and have good suppression of CSF and WM signals. The difference in T1 relaxation times between GM and WM allows suppression of one of these tissues in favor of selectively imaging the other tissue. Because this difference is comparatively small, however, the signal of GM also substantially decreases during TIshort, which leads to a relatively low SNRGM on 3D DIR images. Our current single-slab 3D DIR implementation yields a mean SNRGM of 18.3 for an acquisition time of 9 minutes 47 seconds and a voxel volume of 1.90 mm3. After we took into account differences in voxel volume and acquisition duration, the SNRGM for the single-slab 3D DIR sequence is equivalent to the SNRGM of 30.5 for an interleaved multislab 3D DIR sequence (10). (SNR is proportional to the voxel volume and square root of the total sampling time. For 2.0-mm3 pixels and an acquisition time of 20 minutes, as used for the multislab 3D DIR sequence, the SNR of 18.3 would translate to 18.3 · [2.0 mm3/1.9 mm3] · {surd}[20 minutes/9.78 minutes] = 27.6.)

The regional variations in GM signal intensities on single-slab 3D DIR images are similar to those described on two-dimensional FLAIR images, on which generally the highest signal intensity was in cortical structures in the limbic lobe and the lowest signal intensity was in the cortex along the central sulcus (19). These variations seem to correspond to and may be employed to indicate distinct cortical areas (20,21). In addition, the variation of signal intensities has the same distribution as the T2 relaxation times for a few selected cortical regions (22). Possible regional variations in cortical T1 relaxation times may also play a role, but accurate T1 determinations of the thin and folded cortex require measurements with high in-plane and through-plane spatial resolution. An additional explanation for GM signal intensity differences is that partial volume effects will play a role if the cortex is so thin that the contributing pixels only partly contain GM. These pixels and the corresponding cortical area will then have lower signal intensity. The cortical thickness varies between 1.0 and 4.5 mm, and substantial regional differences have been reported (23). The anterior bank of the central sulcus, however, should be substantially thicker than the posterior bank (2.70 vs 1.76 mm, respectively) according to measurements on inversion-recovery MR images with high spatial resolution and in agreement with postmortem studies (24). Because the single-slab 3D DIR images have low signal intensity on both sides of the central sulcus, the difference in relaxation times probably plays a prominent role.

The homogeneous suppression of WM signals is in agreement with the relatively small range of T1 values of WM, although some regional variation exists (25). T1 values that have been observed by using different techniques at 1.5 T vary considerably (25–28). Therefore, TIs of our sequence have been optimized empirically such that hardly any signal intensity differences can be observed on images of control subjects. An intrinsic limitation of any inversion-recovery technique is the dependence on the actual tissue T1 relaxation time. Interestingly, 3D DIR images of patients with MS have a diffusely higher signal intensity of WM. The suboptimal suppression may indeed be explained by the increased T1 values in so-called normal-appearing WM (WM with apparently normal signal intensity on conventional T2-weighted images) of patients with MS (25).

The duration of the single-slab 3D DIR method (just less than 10 minutes) requires cooperative subjects. Compared with recently published multislab 3D DIR techniques (6,10,11), however, the single-slab 3D DIR method permits a shorter acquisition time for equivalent spatial resolution and completely suppresses flow artifacts. In addition, section-to-section signal intensity and contrast variations along the slab-select direction secondary to an imperfect slab-select profile are eliminated. The TIs of the single-slab 3D DIR sequence were optimized to suppress WM and CSF in control subjects at 1.5 T. Routine clinical imaging can be performed by using these fixed TIs, and whether a suboptimal signal suppression of WM corresponds to a pathologic change of the T1 relaxation time of WM should be further investigated. The use of the single-slab 3D DIR sequence at other field strengths or for other choices of repetition time will require adjustment of the TIs.

Our study had limitations. This study was performed with a limited number of subjects and was meant to show the feasibility of the single-slab 3D DIR MR sequence. At this stage, intrasubject comparisons with techniques such as multislab 3D DIR or two-dimensional DIR have not been performed. Although we have noticed variations in GM signal intensity, in this particular study we have not demonstrated that the single-slab acquisition has a uniform response across the slab. The use of non–spatially selective adiabatic pulses transmitted with the body coil, however, ensures the best possible homogeneity over the entire slab.

The main features of single-slab 3D DIR MR sequences are the high contrast of GM when compared with CSF and WM and the relatively high and nearly isotropic spatial resolution. The combination of high spatial resolution and a clear delineation of the cortex allows a better detection of small intracortical lesions, as well as a proper differentiation between lesion types in patients with MS (6). In addition, the high GM contrast may be employed to obtain a reliable automatic segmentation of the cortex, which is required to determine cortical thickness (5), cortical volume, or cortical abnormalities that may, for instance, identify an epileptic focus. Whether automatic segmentation will benefit from a multichannel analysis of 3D DIR and a 3D data set with another contrast (T1 weighted, T2 weighted, or FLAIR) should be investigated. Furthermore, because of the nearly isotropic resolution, images can be reconstructed in any orientation without image degradation, which facilitates registration in longitudinal studies. A further advantage of the technique is the absence of flow artifacts from blood or CSF. Therefore, the single-slab 3D DIR MR images allow good evaluation of the brainstem and cerebellar areas, as well as of the cortex around the sagittal sinus—a quality that has also been described for single-slab 3D FLAIR images (15).


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
References
 


   FOOTNOTES
 

Abbreviations: CNR = contrast-to-noise ratio • CSF = cerebrospinal fluid • DIR = double inversion recovery • FLAIR = fluid-attenuated inversion recovery • GM = gray matter • MS = multiple sclerosis • SD = standard deviation • SNR = signal-to-noise ratio • 3D = three-dimensional • TI = inversion time • WM = white matter

Authors stated no financial relationship to disclose.

Author contributions: Guarantors of integrity of entire study, P.J.W.P., F.B.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, P.J.W.P., J.P.A.K., J.P.M.; clinical studies, P.J.W.P., F.B.; and manuscript editing, all authors


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 

  1. Miller DH, Barkhof F, Frank JA, Parker GJ, Thompson AJ. Measurement of atrophy in multiple sclerosis: pathological basis, methodological aspects and clinical relevance. Brain 2002;125:1676–1695.[Abstract/Free Full Text]
  2. Scheltens P, Fox N, Barkhof F, De Carli C. Structural magnetic resonance imaging in the practical assessment of dementia: beyond exclusion. Lancet Neurol 2002;1:13–21.[CrossRef][Medline]
  3. Barkovich AJ, Rowley HA, Andermann F. MR in partial epilepsy: value of high-resolution volumetric techniques. AJNR Am J Neuroradiol 1995;16:339–343.[Abstract]
  4. Antel SB, Bernasconi A, Bernasconi N, et al. Computational models of MRI characteristics of focal cortical dysplasia improve lesion detection. Neuroimage 2002;17:1755–1760.[CrossRef][Medline]
  5. Lerch JP, Pruessner JC, Zijdenbos A, Hampel H, Teipel SJ, Evans AC. Focal decline of cortical thickness in Alzheimer's disease identified by computational neuroanatomy. Cereb Cortex 2005;15:995–1001.[Abstract/Free Full Text]
  6. Geurts JJ, Pouwels PJ, Uitdehaag BM, Polman CH, Barkhof F, Castelijns JA. Intracortical lesions in multiple sclerosis: improved detection with 3D double inversion-recovery MR imaging. Radiology 2005;236:254–260.[Abstract/Free Full Text]
  7. Redpath TW, Smith FW. Use of a double inversion recovery pulse sequence to image selectively grey or white brain matter. Br J Radiol 1994;67:1258–1263.[Abstract]
  8. Turetschek K, Wunderbaldinger P, Bankier AA, Zontsich T. Double inversion recovery imaging of the brain: initial experience and comparison with fluid attenuated inversion recovery imaging. Magn Reson Imaging 1998;16:127–135.[CrossRef][Medline]
  9. Bedell BJ, Narayana PA. Implementation and evaluation of a new pulse sequence for rapid acquisition of double inversion recovery images for simultaneous suppression of white matter and CSF. J Magn Reson Imaging 1998;8:544–547.[Medline]
  10. Boulby PA, Symms MR, Barker GJ. Optimized interleaved whole-brain 3D double inversion recovery (DIR) sequence for imaging the neocortex. Magn Reson Med 2004;51:1181–1186.[CrossRef][Medline]
  11. Pouwels PJ, Kuijer JP, Korf ES, Geurts JJ, Barkhof F. 3D double inversion recovery for high resolution gray matter brain imaging [abstr]. In: Proceedings of the 10th Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2002; 1290.
  12. Tan IL, Pouwels PJ, van Schijndel RA, Ader HJ, Manoliu RA, Barkhof F. Isotropic 3D fast FLAIR imaging of the brain in multiple sclerosis patients: initial experience. Eur Radiol 2002;12:559–567.[Medline]
  13. Mugler JP, Bao SM, Mulkern RV, et al. Optimized single-slab three-dimensional spin-echo MR imaging of the brain. Radiology 2000;216:891–899.[Abstract/Free Full Text]
  14. Mugler JP, Kiefer B, Brookeman JR. Three-dimensional T2-weighted imaging of the brain using very long spin-echo trains [abstr]. In: Proceedings of the Eighth Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2000; 687.
  15. Kallmes DF, Hui FK, Mugler JP. Suppression of cerebrospinal fluid and blood flow artifacts in FLAIR MR imaging with a single-slab three-dimensional pulse sequence: initial experience. Radiology 2001;221:251–255.[Abstract/Free Full Text]
  16. Zhao L, Mugler JP, Wei X, et al. A high-resolution clinical whole-brain scan using single-slab three-dimensional T1W, T2W and FLAIR fast spin-echo sequences [abstr]. In: Proceedings of the 10th Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2002; 1294.
  17. Henkelman RM. Measurement of signal intensities in the presence of noise in MR Images. Med Phys 1985;12:232–233.[CrossRef][Medline]
  18. Gudbjartsson H, Patz S. The Rician distribution of noisy MRI data. Magn Reson Med 1995;34:910–914.[Medline]
  19. Hirai T, Korogi Y, Yoshizumi K, Shigematsu Y, Sugahara T, Takahashi M. Limbic lobe of the human brain: evaluation with turbo fluid-attenuated inversion-recovery MR imaging. Radiology 2000;215:470–475.[Abstract/Free Full Text]
  20. Bendersky M, Rugilo C, Kochen S, Schuster G, Sica RE. Magnetic resonance imaging identifies cytoarchitectonic subtypes of the normal human cerebral cortex. J Neurol Sci 2003;211:75–80.[Medline]
  21. Butman JA, Rebmann AJ. Parcellation of cerebral cortex based on FLAIR signal intensity [abstr]. In: Proceedings of the 12th Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2004; 758.
  22. Georgiades CS, Itoh R, Golay X, van Zijl PC, Melhem ER. MR imaging of the human brain at 1.5 T: regional variations in transverse relaxation rates in the cerebral cortex. AJNR Am J Neuroradiol 2001;22:1732–1737.[Abstract/Free Full Text]
  23. Fischl B, Dale AM. Measuring the thickness of the human cerebral cortex from magnetic resonance images. Proc Natl Acad Sci U S A 2000;97:11050–11055.[Abstract/Free Full Text]
  24. Meyer JR, Roychowdhury S, Russell EJ, Callahan C, Gitelman D, Mesulam MM. Location of the central sulcus via cortical thickness of the precentral and postcentral gyri on MR. AJNR Am J Neuroradiol 1996;17:1699–1706.[Abstract]
  25. van Walderveen MA, van Schijndel RA, Pouwels PJ, Polman CH, Barkhof F. Multislice T-1 relaxation time measurements in the brain using IR-EPI: reproducibility, normal values, and histogram analysis in patients with multiple sclerosis. J Magn Reson Imaging 2003;18:656–664.[CrossRef][Medline]
  26. Griffin CM, Chard DT, Parker GJ, Barker GJ, Thompson AJ, Miller DH. The relationship between lesion and normal appearing brain tissue abnormalities in early relapsing remitting multiple sclerosis. J Neurol 2002;249:193–199.[CrossRef][Medline]
  27. Vaithianathar L, Tench CR, Morgan PS, Lin X, Blumhardt LD. White matter T-1 relaxation time histograms and cerebral atrophy in multiple sclerosis. J Neurol Sci 2002;197:45–50.[CrossRef][Medline]
  28. Steinhoff S, Zaitsev M, Zilles K, Shah NJ. Fast T-1 mapping with volume coverage. Magn Reson Med 2001;46:131–140.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 BrainHome page
M. Bendszus, G. Ladewig, L. Jestaedt, B. Misselwitz, L. Solymosi, K. Toyka, and G. Stoll
Gadofluorine M enhancement allows more sensitive detection of inflammatory CNS lesions than T2-w imaging: a quantitative MRI study
Brain, September 1, 2008; 131(9): 2341 - 2352.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
A. H. Poonawalla, K. M. Hasan, R. K. Gupta, C. W. Ahn, F. Nelson, J. S. Wolinsky, and P. A. Narayana
Diffusion-Tensor MR Imaging of Cortical Lesions in Multiple Sclerosis: Initial Findings
Radiology, March 1, 2008; 246(3): 880 - 886.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Neuroradiol.Home page
Y. Duan, P.G. Hildenbrand, M.P. Sampat, D.F. Tate, I. Csapo, B. Moraal, R. Bakshi, F. Barkhof, D.S. Meier, and C.R.G. Guttmann
Segmentation of Subtraction Images for the Measurement of Lesion Change in Multiple Sclerosis
AJNR Am. J. Neuroradiol., February 1, 2008; 29(2): 340 - 346.
[Abstract] [Full Text] [PDF]

Home page
 NeurologyHome page
R. Zivadinov
Can imaging techniques measure neuroprotection and remyelination in multiple sclerosis?
Neurology, May 29, 2007; 68(22_suppl_3): S72 - S82.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2413051182v1
241/3/873    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Pouwels, P. J. W.
Right arrow Articles by Barkhof, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Pouwels, P. J. W.
Right arrow Articles by Barkhof, F.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE