CME. Original Research

Transcription

1 CME JOURNAL OF MAGNETIC RESONANCE IMAGING 000: (2012) Original Research Novel Application of T1-Weighted BLADE Sequences With Fat Suppression Compared to TSE in Contrast-Enhanced T1-Weighted Imaging of the Neck: Cutting-Edge Images? Thomas Finkenzeller, MD, 1,2 * Niels Zorger, MD, 1,3 Thomas Kühnel, MD, 4 Christian Paetzel, MD, 2 Gerhard Schuierer, MD, 5 Christian Stroszczynski, MD, 1 and Claudia Fellner, PhD 1 Purpose: To evaluate if the use of BLADE sequences might overcome some limitations of magnetic resonance imaging (MRI) in the extracranial head and neck, which is a diagnostically challenging area with a variety of artifacts and a broad spectrum of potential lesions. Materials and Methods: After informed consent and Institutional Review Board approval, two different BLADE sequences with (BLADE IR) and without inversion pulse (BLADE) were compared to turbo-spin echo (TSE) with fat saturation for coronal T1-weighted postcontrast imaging of the extracranial head and neck region in 40 individuals of a routine patient collective. Visual evaluation of image sharpness, motion artifacts, vessel pulsation, contrast of anatomic structures, contrast of pathologies to surrounding tissue as well as BLADE-specific artifacts was performed by two experienced, independent readers. Statistical evaluation was done by using the Wilcoxon test. Results: Both BLADE and BLADE IR were significantly superior to TSE regarding pulsation artifacts and delineation of thoracic structures. TSE provided better results concerning contrast muscle/fat tissue and contrast lymph nodes/fat. More important, it showed significantly better contrast of several lesions, facilitating the detection of patient pathology. 1 Institute of Radiology, University Medical Center Regensburg, Regensburg, Germany. 2 Institute of Radiology, Klinikum Weiden, Weiden, Germany. 3 Institute of Radiology, Krankenhaus Barmherzige Brüder Regensburg, Regensburg, Germany. 4 Department of Otorhinolaryngology, University Medical Center Regensburg, Regensburg, Germany. 5 Center of Neuroradiology, University Medical Center and Bezirksklinikum Regensburg, Regensburg, Germany. Presented at the 2010 annual RSNA meeting Chicago/USA SST *Address reprint requests to: T.F., Institute of Radiology and Neuroradiology, Klinikum Nuernberg Sued, Breslauerstrasse 201, Nuernberg, Germany. thomas.finkenzeller@klinikum-nuernberg.de Received September 11, 2011; Accepted August 29, DOI /jmri View this article online at wileyonlinelibrary.com. Conclusion: T1-weighted coronal imaging of the extracranial head and neck region is demanding. T1-weighted BLADE sequences still have drawbacks in anatomical contrast and lesion detection but offer possibilities to achieve reasonable image quality in difficult cases with a variety of artifacts. Key Words: MRI; neck; BLADE; artifacts; PROPELLER J. Magn. Reson. Imaging 2012;000: VC 2012 Wiley Periodicals, Inc. MAGNETIC RESONANCE IMAGING (MRI) of the head and neck region is demanding for several reasons. Most of the anatomical structures in this area are rather small and adjacent to a variety of moving or movable structures. The change in contour between the head and the trunk causes inhomogeneities in the magnetic field that degrades image quality (1) at the head-neck-shoulder junction and makes this region vulnerable to susceptibility artifacts. Furthermore, a broad spectrum of artifacts including pulsation and swallowing artifacts may occur even in cooperative patients (2,3). Also, the wide spectrum of potential mass lesions in this area, ranging from tumors, metastases, pathologic lymph nodes, and thyroid nodules to vascular malformations is demanding and favors imaging modalities that provide a high contrast and a fair resolution of anatomical structures. Assessing the extent of lesions and detecting metastatic spread is the major objective of modern imaging and therefore demands high standards, especially of MRI with its superior soft tissue contrast. Nowadays, periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) as well as the Siemens Medical Solutions implementation called BLADE sequence are used to reduce artifacts induced by in-plane rotation and translational motion using an alternative way of k-space sampling. The underlying basic principle of this technique has been described by Pipe (4) in Unlike rectilinear VC 2012 Wiley Periodicals, Inc. 1

2 2 Finkenzeller et al. k-space sampling, PROPELLER or BLADE acquires multiple echo trains which cover the k-space in rotating and partially overlapping so-called blades. Clinical applications showed the benefit of these sequences in MRI of the brain to reduce motion artifacts in uncooperative or pediatric patients (5,6) or to suppress flow artifacts (7). In abdominal imaging they can reduce artifacts of bowel movement (8,9). In nearly all reported applications PROPELLER or BLADE have been used in transverse orientation where the rotating field of view (FOV) is not a major risk to induce foldover artifacts in the phase-encoding direction. Most publications focused on T2-weighted fast spin-echo (FSE) or FSE-based diffusion-weighted techniques (10,11), whereas only a minority of articles described T1-weighted applications. Up to now (PubMed research, Feb. 2012), no publications on the use of T1-weighted imaging with fat saturation are available. Therefore, the aim of our study was to apply and evaluate BLADE for coronal postcontrast T1-weighted imaging with fat saturation of the neck region and to evaluate it in a routine clinical setting. For this purpose, image quality, contrast of relevant anatomical structures, and the presence and extent of various artifacts were evaluated in 40 consecutive patients using an optimized coronal T1-weighted turbo-spin echo (TSE) sequence with spectral fat saturation and two different BLADE sequences with (BLADE IR) and without (BLADE) inversion pulse. MATERIALS AND METHODS Patients Forty consecutive patients, 27 men and 13 women (age range: years, mean: years), referred for MRI of the neck, were included in this prospective study. All were ENT patients in treatment for different pathologies and all showed pathological findings in the MRI scan. The study was approved by the Institutional Review Board and all patients provided written informed consent. The final MR findings in all patients were judged from the complete MR examination including standard transverse, coronal, and if necessary sagittal images. We found a variety of neck lesions including the top three lymphadenopathy (n ¼ 9), tongue carcinoma (n ¼ 6), and thyroid lesions (n ¼ 6). Furthermore, there were four pharyngeal carcinomas and three tongue transplants after curative tumor operation. The remaining pathologies were hemangiomas, soft tissue metastasis, lymphomas, bony metastasis in the mandible, and laryngeal cancer (all, n ¼ 2). In addition, one clival chordoma and one parotid carcinoma were detected. MR Examination MRI of the neck region was performed at 1.5 T (Magnetom Avanto, Siemens Healthcare, Erlangen, Germany) using a combination of head, neck, and spine array coils. The scanner has a gradient system Table 1 Measurement Parameters for Study Sequences : Coronal T1-Weighted TSE, BLADE, and BLADE IR TSE BLADE BLADE IR TR [msec]/te 607/10 607/ /56 [msec] TI [msec] 860 Echo train length (ETL) Bandwidth [Hz/pixel] (BW) Slice thickness [mm] (SD) Field of view [mm mm] (FOV) Matrix size (MA) Acquisition time [min:s] (AT) 3:27 4:09 4:34 with 45 mt/m maximum gradient field strength and 200 T/m/s slew rate. After administration of gadolinium-dtpa (0.1 mmol/kg) coronal T1-weighted TSE was acquired in all patients as the first postcontrast sequence followed by randomized acquisition of the two T1-weighted BLADE sequences. None of the sequences were repeated even if image quality was insufficient due to motion or other artifacts. Table 1 shows the parameters used in our study sequences. For T1-weighted TSE imaging we applied our routine TSE sequence with fat saturation and TR 607 / TE 10, echo-train-length (ETL) 3, imaging matrix (MA) of and 3.27 minutes acquisition time (TA) with flow compensation in the readout direction to reduce flow artifacts (Table 1). Both BLADE sequences were matched to the aforementioned sequence regarding FOV and slice thickness (SD). One of the BLADE sequences used an inversion pulse at 860 msec (BLADE IR) (Table 1). Image reconstruction of the BLADE acquisitions was performed without inplane motion correction because ETL of the BLADE sequence is too short in the current implementation of the sequence. For BLADE IR in-plane motion correction would have been possible, but was abandoned for the purposes of this study and planned for use in further studies. Besides the comparison for T1-weighted TSE coronal imaging with the adapted BLADE and BLADE IR sequence, coronal T1-weighted precontrast (TR: 442; TE 12; SD 5 mm; FOV 300; MA ) and STIR images (5520; 73; TI 150; 5 mm; 300; ) as well as slices in transverse orientation T2-weighted (3830; 85; 5 mm; 250; ) and T1-weighted images before (470; 12; 5 mm; 250; ) and after (554; 17; 5 mm; 250; ) contrast application were acquired in all patients. Depending on the pathology, additional images in sagittal orientation (462; 17; 4 mm; 300; ) were added if necessary. In order to assess the spatial resolution of TSE and BLADE sequences we used a resolution phantom similar to the phantom described by Fellner et al (12), which allows a simple visual evaluation of spatial

3 CE-T1-Weighted BLADE Imaging of the Neck 3 resolution. The phantom contains parallel Plexiglas strips from 0.1 to 1.5 mm thickness. They are fixed in pairs of five with the distance of the strips being equal to their thickness. This phantom proved to be well suited for the visual assessment of spatial resolution of different MRI techniques in former studies (12). To compare the intrinsic resolution of evaluated sequences, all of them were applied with an FOV of 300 mm and a matrix size of The measurement parameters influencing contrast behavior were identical to those used in the patients protocol. Image Evaluation Visual assessment of image quality was performed by two independent readers blinded to the imaging technique as well as to patient data, medical history, and additional MR images. Reader 1 was an experienced neuroradiologist, Reader 2 an experienced radiologist, both with more than 10 years experience in MRI. Image quality was graded on a scale from 1 to 5 (1: excellent, without any impairment of image quality 2: good, with only minimal impairment of image quality 3: moderate, showing artifacts that diminish image quality but are still diagnostic 4: poor, with severe impairment of image quality and limited diagnostic reliability 5: nondiagnostic, artifacts/alterations are too severe to make a diagnosis) for the following criteria: image sharpness, overall motion artifacts, pulsation artifacts related to vessels, visualization of thoracic and pulmonary structures, contrast muscle/fat tissue, contrast lymph nodes/fat, and contrast lesion/surrounding tissue (Table 2). Additional image data as well as follow-up examinations and/or histological work-up (if available) were included to obtain the final diagnosis after evaluation of the datasets. Statistical Analysis All statistical calculations and tests were performed using SPSS software (v. 17.0, Chicago, IL). Results of the visual evaluation for TSE and BLADE sequences were compared with the two-sided Wilcoxon rank sum test for each individual reader as well as for the mean grading of both readers. We looked for interobserver agreement of the visual evaluation using the Pearson correlation coefficient (Table 3). For all tests P < 0.05 were considered statistically significant. RESULTS Evaluation of the comparability of the study sequences regarding image resolution was performed using a suitable resolution phantom. In this in vitro setup, visual evaluation of the phantom performed by an experienced MRI physicist (C.F.) revealed identical resolution of the BLADE and the BLADE IR sequence compared to TSE (Fig. 1a c): Using a pixel size of 0.78 mm, the 0.8-mm pattern was clearly resolved in all sequences, whereas the 0.75 mm pattern showed some limitations concerning the delineation of the Table 2 Results of the Visual Evaluation on a Scale From 1 (Excellent) to 5 (Nondiagnostic): Means and Standard Deviation Reader 1 Reader 2 Mean (reader 1, reader 2) TSE BLADE BLADE-IR TSE BLADE BLADE-IR TSE BLADE BLADE-IR Sequence Anatomical structures and artifacts Image sharpness * Motion artifacts * Pulsation artifacts *** *** *** *** *** *** BLADE typical artifacts *** *** *** *** *** *** Delineation of thoracic *** *** ** ** *** *** structures Muscle/fat contrast * Lymph nodes/fat contrast * *** * *** ** *** Pathologies (overall and 3 most frequent) Overall Lesion contrast ** ** *** *** *** *** Pathologic lymph nodes (n ¼ 9) Tongue carcinoma (n ¼ 6) Thyroid Lesions (n ¼ 6) Wilcoxon rank sum test, significant difference between TSE and BLADE: *P < 0.05; **P < 0.01; ***P <

4 4 Finkenzeller et al. Table 3 Interobserver Agreement of the Visual Assessment Using Pearson Correlation Coefficient TSE BLADE BLADE IR Image sharpness 0.40* 0.43** 0.07 Motion artifacts ** Pulsation artifacts BLADE typical artifacts Delineation of thoracic 0.64** 0.59** 0.55** structures Muscle/fat contrast * Lymph nodes/fat contrast ** 0.36* Lesion contrast 0.70** 0.62** 0.83** *P < 0.05; **P < 0.01; ***P < individual strips. This means that the intrinsic resolution of both BLADE sequences is identical to the resolution of TSE in a nonmoving object. In the patient study we visually compared the study sequences TSE, BLADE, and BLADE IR concerning several criteria. Both BLADE and BLADE IR were superior to TSE regarding pulsation artifacts related to vessels and the delineation of thoracic structures (Fig. 2a c). The difference was statistically significant for both readers as well as for their mean grading. Both adapted sequences (BLADE and BLADE IR) showed similar results compared to TSE concerning contrast muscle/fat tissue. Contrast lymph nodes/fat (Fig. 3a c) and the overall grading of lesion/surrounding tissue was significantly better for TSE for each reader and the mean grading for both readers as well. The number of examinations judged as nondiagnostic by at least one reader was very heterogeneous. TSE images were judged nondiagnostic due to vessel pulsation artifacts in eight patients, while both BLADE sequences were diagnostic in all of the patients regarding this criterion. The delineation of thoracic structures was graded nondiagnostic in 17 TSE but only four of the BLADE and five of the BLADE IR examinations. In one BLADE and one BLADE IR examination the images showed so-called BLADE-typical artifacts (Fig. 4b,c), leading to a nondiagnostic grading in the area of these star-like artifacts (at the level of the glottis). None of the TSE sequences did show these artifacts. BLADE IR missed one carcinoma on the lateral tongue margin which was not detected by either reader in this sequence (Fig. 4a c). The remaining pathological entities were detected with each of the three sequences applied. Regarding all 40 pathologies evaluated, TSE showed a significantly better contrast between lesion and surrounding tissue in the mean grading of both readers compared to both contending BLADE sequences. Lesion contrast was scored poor, but still diagnostic (grade 4) for only two and three cases, respectively (Reader 1; Reader 2) of the TSE- Figure 1. Examples from the resolution phantom, which contained parallel Plexiglas strips from 0.1 to 1.5 mm thickness and equal distance. Although truncation artifacts parallel to the patterns were present, they did not influence the visual evaluation of spatial resolution. It showed similar spatial resolution of both BLADE sequences (b,c) compared to TSE (a). The 0.8-mm pattern (arrow) could be clearly resolved in all sequences using a pixel size of 0.78 mm. The 0.75-mm pattern (arrowhead) showed limitations concerning the delineation of the individual strips.

5 CE-T1-Weighted BLADE Imaging of the Neck 5 Figure 2. In the sequences with rotating k-space sampling (BLADE (b) and BLADE IR (c)) differentiation of mediastinal and pulmonary as well as vascular structures was clearly easier than in TSE (a). sequences. For BLADE IR both readers judged seven examinations and for the BLADE sequence one and four, respectively, patients as grade 4. Comparing the three sequences for detection of the different pathologies per lesion type there was no statistical evaluation performed due to the small number of cases. For the most frequent lesions, a positive trend for the TSE sequence was seen in pathologic lymph nodes and imaging of tongue carcinoma. Only the evaluation of lesions within the thyroid gland was judged slightly superior for the BLADE IR sequence compared to TSE (Table 2). For interobserver agreement of the visual evaluation using the Pearson correlation coefficient (Table 3), there was only fair to moderate agreement between the readers concerning image sharpness, motion artifacts, vessel pulsation, and contrast between muscles and fat as well as lymph nodes and fat. Quite good agreement was reached for the delineation of thoracic structures. For lesion detection and judgment the interobserver agreement was 0.70 for TSE, 0.62 for BLADE, and 0.83 for BLADE IR, resulting in a good consensus in this important part of the evaluation. DISCUSSION PROPELLER or BLADE sequences have proven to be superior to standard rectilinear data sampling with regard to various imaging artifacts in several studies (12,13). Their k-space acquisition scheme with rotating, partially overlapping, so-called blades is beneficial in noncooperating patients causing motion artifacts during data sampling (5,6). Extracranial head and neck region with its complex anatomy, its broad spectrum of potential pathologies, and large variety of region immanent artifacts poses significant problems for all imaging modalities. Sufficient contrast of quite small anatomical structures and sharp images free of interfering artifacts are important requirements of medical imaging in this diagnostically challenging region. MRI-based detection and diagnostic workup of

6 6 Finkenzeller et al. Figure 3. Visually evaluating contrast of muscle/fat and (pathologic and nonpathologic) lymph nodes/fat (arrow) as well as lesion contrast to surrounding tissue, BLADE (b) and BLADE IR (c) images were significantly inferior to TSE (a) in T1- weighted postcontrast coronal plane. Especially BLADE IR showed lower contrast compared to TSE for these criteria. neoplasms and/or benign lesions depends on the distinction between the contrast of the lesion and that of adjacent tissue in T1- and T2-weighted images. Furthermore, the high signal of surrounding fat potentially masks the contrast enhancement of lesions in T1-weighted images and also obscures them in T2- weighted sequences. The benefit of gadolinium as a contrast agent in T1-weighted MRI is well documented (1,13), but without fat suppression techniques its value is limited in the head and neck region. Therefore, fat-saturated T1-weighted images after contrast application are an essential part of MRI of the neck region and are usually acquired in coronal and transverse plane. To our knowledge only very few applications of BLADE (or PROPELLER)-based T1-weighted techniques in patients have been reported up to today (7,14,15) and most evaluations have been limited to axial slice orientations (16). One recent publication on T2-weighted BLADE imaging of the neck showed artifact reduction and improvement of image quality compared to T2-weighted FSE sequences with equivalent lesion detection (17). To assess the potential role of two different T1- weighted BLADE techniques with fat saturation in comparison to the standard TSE technique with spectral fat saturation, contrast of relevant anatomical structures and tissue was studied qualitatively by visual evaluation. In conventional TSE sequences sufficient T1 contrast is achieved by a very short TE and short ETL. A major drawback of the BLADE/PROPELLER sequence in T1-weighted imaging is the need for a rather long TE, which is contrary to the desired T1-weighting. One way to obtain reasonable T1 contrast is the use of a TE as short as possible (achieved by the use of an extremely high readout bandwidth), which in our BLADE sequence was set at 23 msec. Nevertheless,

7 CE-T1-Weighted BLADE Imaging of the Neck 7 Figure 4. In lesion detection BLADE IR (c) missed one carcinoma of the left lateral tongue margin in this patient (arrow). Images (a c) were acquired in the displayed order and visualize a decline in lesion contrast in T1-weighted postcontrast coronal plane. This is not necessarily a drawback of the sequence used but could also be attributed to contrast medium washout. Additionally, in images (b,c) star-like-artifacts at the level of the glottis can be seen (arrowhead), which are typical of BLADE sequences, but rarely alter the image quality significantly. this is still a quite long TE, resulting in a significant loss of T1 contrast that might explain the inferiority of the BLADE sequence for lesion contrast and the differentiation of lymph nodules to fat. A second possibility of achieving sufficient T1 contrast is the application of an inversion pulse, as we did in our inversion recovery sequence (BLADE IR). This allowed the use of a longer TE and therefore a prolongation of the ETL and a reduction of the bandwidth. In the visual evaluation of our routine patient collective, coronal BLADE and BLADE IR were significantly superior to TSE with regard to vessel related pulsation artifacts. The overlapping and repeated measurement of central k-space areas seems to enable sufficient correction of pulsation artifacts of carotid and vertebral arteries as well as flow phenomena and venous superimposition (13) with the BLADE acquisition scheme. As a side-effect of the rotating blades, star-like artifacts may arise in some areas of high contrast structures, as, for example, the epiglottic transition to the pharyngeal space. Similar artifacts are known in transverse BLADE imaging of the brain at 3T arising in the 4th ventricle at the inflow region of the cerebral aqueduct (14,15). Usually, these BLADE-typical artifacts are easy to detect and have no major impact on diagnostic reliability. Furthermore, BLADE-specific wraparound artifacts appear at the lower corners of the images (13), but do not alter image quality to a noticeable degree (18). As most imaging studies of the extracranial head and neck also cover the upper part of the mediastinum and the pulmonary apex, we also evaluated delineation of aortic arch and supra-aortic vessels, pulmonary vessels, and pulmonary tissue as well as possible lesions within the depicted lung sections.

8 8 Finkenzeller et al. Both readers preferred BLADE images in the assessment of this particular area as diagnostic reliability of TSE images was severely impaired by mainly pulsation and breathing artifacts (Table 2). Regarding gross motion artifacts our evaluation of the neck region showed no significant difference of both BLADE sequences compared to TSE, which was not consistent with a study concerning BLADE imaging of the cervical spine (13). While the partially overlapping rotating blades scheme seems to be helpful in reducing minor local artifacts, gross motion is only compensated to a markedly lesser extent by this form of k-space sampling. Intrinsic motion correction caused by the altered k-space coverage in BLADE sequences is optimal for very long echo trains. Therefore, using shorter echo trains for both T1-weighted BLADE sequences degrades the intrinsic motion correction. For this purpose, the additional motion correction algorithm (which was not used in our study) might yield better results. The matrix size of our routine TSE sequence ( ) could not be realized with BLADE sequences in the current implementation (which is limited to a maximum matrix size of ), which might explain similar or even inferior image sharpness of BLADE compared with TSE. The phantom measurements demonstrated equivalent resolution and image sharpness when using identical geometric parameters. In BLADE IR similar imaging parameters (TR, TE, ETL) proved to be useful in former studies (7) regarding T1-contrast in brain imaging. In the BLADE sequence without inversion pulse, the chosen parameters for BLADE TE, TR, and ETL are an extreme compromise not used until now for T1-weighted BLADE imaging. But in combination with fat saturation necessary for T1-weighted neck imaging, these parameters need to be optimized to enable sufficient lesion contrast. When looking at all evaluated pathologies in our study TSE showed significantly better results compared to both contending BLADE sequences in the detection of pathologic lesions including lymphadenopathy, different types of oral and pharyngeal cancer, as well as some rare ENT entities. In the overall judgment of both readers, TSE was preferred for lesion detection, especially concerning pathologic lymph nodes and oral cavity cancer. BLADE IR even masked one carcinoma at the lateral tongue margin that was clearly detected by BLADE and TSE. Only for the diagnosis of pathologies within the thyroid gland did BLADE sequences show advantages (BLADE IR) or were at least equal to TSE (without significance). But as MRI is not the optimal and first imaging modality for thyroid lesions, this benefit is rather minor concerning lesion detection and assessment. The design of our study has two major limitations: Regarding sequence parameters, the matrix size of TSE is higher than in the BLADE sequence, which potentially limits the geometrical comparability at least in the patient study and also could lower image sharpness in BLADE. As the phantom study in a nonmoving object showed no visually detectable difference of the study sequences compared to TSE, we would suggest that the mismatch in matrix size has no or only minor implications concerning these criteria. Furthermore, randomized sequence acquisition with alternating order of TSE, BLADE, and BLADE IR would be optimal to overcome the drawbacks of contrast medium washout and increasing patient motion with prolonged examination duration. As we evaluated our patients in a clinical setting we used TSE as firstline imaging to ensure standard validated clinical imaging. This resulted in acquisition of the BLADE sequences 8 to 12 minutes after application of the contrast medium. Therefore, contrast washout with consecutive loss of lesion contrast and potentially increasing patient motion might be reasons for diminished lesion detection of the BLADE sequences in our study. Further studies with randomized sequence acquisition order are necessary to overcome these limitations. Interobserver agreement of the visual evaluation in this study (Table 3) was only fair to moderate for most evaluated items including image sharpness, motion artifacts, vessel pulsation, and contrast between muscles and fat as well as lymph nodes and fat. This might be due to the fact that the readers judged the given criteria over the complete examination of the neck, selecting any image slice and region they liked for judgment of the particular criterion. Therefore, the judgment of, for example, pulsation or other artifacts is very subjective and individual and may be the reason for a rather bad interobserver agreement. But for lesion detection, as the most important part of image evaluation, the interobserver agreement was good for all three sequences, which shows that, if focused on the same point and pathology, the examined BLADE sequences seem to be at least equal compared to TSE. Lesion detection is the main objective of head and neck imaging and therefore missing a pathology due to modality-related limitations or sequence parameters is not acceptable. As T1-weighted MRI in coronal and transverse orientation is a mainstay of modern neck imaging, further efforts have to be made to overcome the problems standard T1-weighted TSE imaging still has in this diagnostically challenging region. In the current application both BLADE sequences showed drawbacks mainly concerning T1 contrast without any clear advantage in favor of BLADE or BLADE IR. But both offer the possibility to reduce several artifacts and therefore their use might improve image quality in selected patients who are not able to cooperate sufficiently or show significant artifacts that are beyond patient influence (eg, vessel pulsation). In conclusion, currently the coronal T1-weighted BLADE sequences evaluated in this study cannot replace standard postcontrast T1-weighted TSE imaging with fat saturation, but in difficult cases with severe pulsation and swallowing artifacts they may be a helpful adjunct to achieve reasonable image quality. REFERENCES 1. Ross MR, Schomer DF, Chappell, Enzmann DR. MR imaging of head and neck tumors: comparison of T1-weighted contrastenhanced fat-suppressed images with conventional T2-weighted and fast spin-echo T2-weighted images. AJR 1994;163:

9 CE-T1-Weighted BLADE Imaging of the Neck 9 2. Melhem ER. Technical challenges in MR imaging of the cervical spine and cord. Magn Reson Imaging Clin N Am 2000;8: Taber KH, Herrick RC, Weathers SW, Kumar AJ, Schomer DF, Hayman LA. Pitfalls and artifacts encountered in clinical MR imaging of the spine. Radiographics 1998;18: Pipe JG. Motion correction with PROPELLER MRI: application to head motion and free-breathing cardiac imaging. Magn Reson Med 1999;42: Forbes KP, Pipe JG, Bird CR, et al. PROPELLER MRI. Clinical testing of a novel technique for quantification and compensation of head movement. J Magn Reson Imaging 2001;14: Forbes KP, Pipe JG, Karis JP, Farthing V, Heiserman JE. Brain imaging in the unsedated pediatric patient: comparison of periodically rotated overlapping parallel lines with enhanced reconstruction with single shot fast spin-echo sequences. AJNR Am J Neuroradiol 2003;24: Naganawa S, Satake H, Iwano S, et al. Contrast-enhanced MR imaging of the brain using T1-weighted FLAIR with BLADE compared with a conventional spin-echo sequence. Eur Radiol 2008; 18: Hirokawa Y, Isoda H, Maetani YS, et al. MRI artifact reduction and quality improvement in the upper abdomen with PROPEL- LER and prospective acquisition correction (PACE) technique. AJR Am J Roentgenol 2008;191: Hirokawa Y, Isoda H, Maetani YS, Arizono S, Shimada K, Togashi K. Evaluation of motion correction effect and image quality with the periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) (BLADE) and parallel imaging acquisition technique in the upper abdomen. J Magn Reson Imaging 2008;28: Forbes KP, Pipe JG, Karis JP, Heiserman JE. Improved image quality and detection of acute cerebral infarction with PROPELLER diffusion-weighted MR imaging. Radiology 2002;225: Pipe JG, Farthing VG, Forbes KP. Multishot diffusion-weighted FSE using PROPELLER MRI. Magn Reson Med 2002;47: Fellner C, Mueller W, Georgi J, Taubenreuther U, Fellner FA, Kalender WA. A high resolution phantom for MRI. Magn Reson Imaging 2001;19: Fellner C, Menzel C, Fellner FA, et al. BLADE in sagittal T2- weighted MR imaging of the cervical spine. Am J Neuroradiol 2010;31: Wintersperger BJ, Runge VM, Biswas, et al. Brain magnetic resonance imaging at 3 Tesla using BLADE compared with standard rectilinear data sampling. Invest Radiol 2006;41: Attenberger UI, Runge VM, Williams KD, et al. T1-weighted brain imaging with a 32-channel coil at 3T using TurboFLASH BLADE compared with standard cartesian k-space sampling. Invest Radiol 2009;44: Alkan O, Kizilkilic O, Yildirim T, Alibek S. Comparison of contrast-enhanced T1-weighted FLAIR with BLADE, and spin-echo T1-weighted sequences in intracranial MRI. Diagn Interv Radiol 2009;15: Ohgiya Y, Suyama J, Seino N, et al. MRI of the neck at 3T using the periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) (BLADE) sequence compared with T2-weighted fast spin-echo sequences. J Magn Reson Imaging 2010;32: Michaely HJ, Kramer H, Weckbach S, Dietrich O, Reiser MF, Schoenberg SO. Renal T2-weighted turbo-spin-echo imaging with BLADE at 3.0 Tesla: initial experience. J Magn Reson Imaging 2008;27: