PROPELLER Technique to Improve Image Quality of MRI of the Shoulder

Transcription

1 Musculoskeletal Imaging Original Research Dietrich et al. PROPELLER Technique in MRI of Shoulder Musculoskeletal Imaging Original Research Tobias J. Dietrich 1 Erika J. Ulbrich 1 Marco Zanetti 1 Sandro F. Fucentese 2 Christian W. A. Pfirrmann 1 Dietrich TJ, Ulbrich EJ, Zanetti M, Fucentese SF, Pfirrmann CWA Keywords: arthrography, imaging sequences, MRI, shoulder, technology assessment DOI: /AJR Received October 24, 2010; accepted after revision May 4, Department of Radiology, Orthopedic University Hospital Balgrist, Forchstrasse 340, Zurich CH-8008, Switzerland. Address correspondence to T. J. Dietrich (tdiet@gmx.net). 2 Department of Orthopedic Surgery, Orthopedic University Hospital Balgrist, Zurich, Switzerland. WEB This is a Web exclusive article. AJR 2011; 197:W1093 W X/11/1976 W1093 American Roentgen Ray Society PROPELLER Technique to Improve Image Quality of MRI of the Shoulder OBJECTIVE. The purpose of this article is to evaluate the use of the periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) technique for artifact reduction and overall image quality improvement for intermediate-weighted and T2-weighted MRI of the shoulder. SUBJECTS AND METHODS. One hundred eleven patients undergoing MR arthrography of the shoulder were included. A coronal oblique intermediate-weighted turbo spinecho (TSE) sequence with fat suppression and a sagittal oblique T2-weighted TSE sequence with fat suppression were obtained without (standard) and with the PROPELLER technique. Scanning time increased from 3 minutes 17 seconds to 4 minutes 17 seconds (coronal oblique plane) and from 2 minutes 52 seconds to 4 minutes 10 seconds (sagittal oblique) using PRO- PELLER. Two radiologists graded image artifacts, overall image quality, and delineation of several anatomic structures on a 5-point scale (5, no artifact, optimal diagnostic quality; and 1, severe artifacts, diagnostically not usable). The Wilcoxon signed rank test was used to compare the data of the standard and PROPELLER images. RESULTS. Motion artifacts were significantly reduced in PROPELLER images (p < 0.001). Observer 1 rated motion artifacts with diagnostic impairment in one patient on coronal oblique PROPELLER images compared with 33 patients on standard images. Ratings for the sequences with PROPELLER were significantly better for overall image quality (p < 0.001). Observer 1 noted an overall image quality with diagnostic impairment in nine patients on sagittal oblique PROPELLER images compared with 23 patients on standard MRI. CONCLUSION. The PROPELLER technique for MRI of the shoulder reduces the number of sequences with diagnostic impairment as a result of motion artifacts and increases image quality compared with standard TSE sequences. PROPELLER sequences increase the acquisition time. T he periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) technique reduces motion artifacts in MRI of the abdomen, chest, spine, and brain [1 7]. Motion artifact reduction with PROPELLER is especially useful in noncooperative patients, such as pediatric patients and those with stroke. In MRI of the musculoskeletal system, motion artifacts are often less pronounced. However, especially in the shoulder, motion artifacts because of breathing or uncooperative patients may have an important influence on the image quality and the diagnostic value. Acronyms used by manufacturers for sequences analogous to the PRO- PELLER technique include BLADE, JET, MultiVane, PROPELLER, and RADAR. Standard turbo spin-echo (TSE) sequences fill the k-space in a rectilinear pattern dur- ing a TR period. Therefore, the data in the center of the k-space are acquired once or according to the number of multiple averages. However, data in the center of the k-space have important influences on the image characteristics, and object motions severely affect the image quality. The PROPELLER sequence is a radial k-space sampling concept with parallel data lines rotating around the center of the k-space, which allows correcting spatial inconsistencies. Data indicating through-plane motion based on correlation measurement are rejected. Motion artifacts are further reduced through averaging in low spatial frequencies [1]. The purpose of this study was to evaluate the use of the PROPELLER technique for intermediate- and T2-weighted MRI of the shoulder. We hypothesized that the PRO- PELLER sequence would reduce motion ar- AJR:197, December 2011 W1093

2 Dietrich et al. tifacts and improve overall image quality, even if it prolonged the lengths of the scans and accentuated susceptibility artifacts. Subjects and Methods Patients One hundred eleven consecutive patients (47 women and 64 men) undergoing MR arthrography of the shoulder were consecutively included. The indications for MR arthrography were suspected rotator cuff tear (n = 48), rotator cuff tear after surgical repair (n = 10), impingement (n = 9), glenohumeral instability (n = 8), nonspecific pain (n = 7), frozen shoulder (n = 7), biceps tendon tear (n = 6), lesion of the biceps tendon anchor (n = 5), degenerative osteoarthritis of the acromioclavicular joint (n = 4) and of the glenohumeral joint (n = 1), calcific tendinitis (n = 3), radiographically occult fracture (n = 2), and pulley lesion (n = 1). Patients were referred by orthopedic surgeon (n = 52), rheumatologist (n = 28), trauma surgeon (n = 15), family physician (n = 9), internist (n = 4), and chiropractor (n = 3). Images were acquired from 71 right and 40 left shoulders. Seventeen shoulders had surgery before MRI and metallic devices, such as anchors, were implanted in 15 shoulders. The mean patient age was 51.3 years (range, years). All 111 patients were consecutively included during a period from February until May All patients in this study gave written permission for anonymized use of their medical data for scientific purposes before the imaging examination. The study was submitted to the institutional review board, and a waiver for additional approval was issued for this study. Imaging Protocol MR arthrography of the shoulder is the standard technique to assess internal derangements of the glenohumeral joint for most patients at our institution. Before MRI, 8 12 ml of 2 mmol/l gadopentetate dimeglumine solution (Magnevist, Bayer Pharma) was injected into the glenohumeral joint under fluoroscopic guidance. The patients were examined in one of two different 1.5-T MRI units (Magnetom Espree and Magnetom Avanto, Siemens Healthcare), depending on scanner availability. Two sequences out of the standard protocol were acquired with and without PROPELLER: a coronal oblique intermediateweighted TSE MRI with fat suppression and a sagittal oblique T2-weighted TSE MRI with fat suppression. The commercial implementation of the PROPELLER technique of the Siemens Healthcare MRI scanner is called BLADE and was used in this study. MRI resolution (FOV, mm; matrix, ; section thickness, 4.0 mm) and the number of signals acquired (n = 1) were the same for both techniques and both MRI scanners. The sequence parameters varied slightly and are given in Table 1. The other three sequences that are included in the standard protocol were not evaluated in this investigation: coronal T1-weighted fast spin-echo images with fat suppression (Avanto: TR/TE, 667/12; FOV, mm; matrix, ; section thickness, 3.0 mm), sagittal T1-weighted spin-echo images (Avanto: TR/TE, 450/12; FOV, mm; matrix, ; section thickness, 4.0 mm), and transverse 3D water-excitation true fast imaging with steady-state precession MRI (Avanto: TR/TE, 11.98/5.15; FOV, mm; matrix, ; section thickness, 1.7 mm; flip angle, 28 ). The standard coronal oblique and sagittal oblique sequences were acquired before the corresponding PROPELLER images in the first 55 shoulders, and PROPELLER images were obtained before the corresponding standard sequence in the remaining patients examined. MRI Evaluation Two board-certified musculoskeletal radiologists evaluated the MRI examinations separately after a training session on 10 cases not included in the analysis. They were unaware of the clinical data. The readers were also unaware of the specific sequence information (i.e., imaging parameters and standard vs PROPELLER technique). The following parameters were analyzed: motion artifacts (n = 111), susceptibility artifacts (n = 111), pulsation artifacts (n = 111), overall image quality (n = 111), and delineation of several anatomic TABLE 1: Measurement Parameters for MRI Examinations Sequence, Parameters structures with respect to the diagnostic value of an MR arthrography examination as in daily practice, including articular and bursal surface of the supraspinatus tendon (n = 104), biceps anchor (n = 95), glenohumeral cartilage (n = 111), biceps tendon in the rotator cuff interval (n = 95), and the acromioclavicular joint (n = 111). The parameter delineation of articular and bursal surface of the supraspinatus tendon were not determined in patients with severe tendon retraction resulting from fullthickness tear of the supraspinatus tendon (n = 7). There was no evaluation of the biceps anchor (n = 16) or the biceps tendon in the rotator cuff interval (n = 16) in patients with tendon tear or after tenotomy. A classification system of five grades was used: 5, no artifact or optimal diagnostic quality; 4, minimal artifacts present or no diagnostic impairment; 3, minor artifacts or minor diagnostic impairment, details and subtle findings are not assessable (e.g., a superior labral anteroposterior lesion); 2, moderate artifacts or diagnostic possibilities delimited to major findings (e.g., a large full-thickness tear); and 1, severe artifacts or diagnostically not usable. An MRI sequence with motion artifacts of grade 3 or lower would typically be repeated at our institution. Statistical Analysis Values are expressed as median and mean. The number of patients with a score of grade 3 or lower is noted for artifacts and overall image quality. The Wilcoxon signed rank test was used to compare the data of the standard and PROPELLER MRI of each observer. A probability for an α error p less than 0.05 was considered sufficient to reject the zero hypothesis of no difference between standard TSE Magnetom Avanto Scanner, Technique Magnetom Espree Standard PROPELLER Standard PROPELLER Coronal oblique TR/TE 2900/ / / /36 Echo-train length Inter-echo spacing (ms) Receiver bandwidth (Hz/pixel) Scanning time (min:s) 3:17 4:17 3:24 3:52 Sagittal oblique TR/TE 3500/ / / /79 Echo-train length Inter-echo spacing (ms) Receiver bandwidth (Hz/pixel) Scanning time (min:s) 2:52 4:10 3:35 4:10 Note Magnetom Avanto and Magnetom Espree scanners are both from Siemens Healthcare. PROPELLER = periodically rotated overlapping parallel lines with enhanced reconstruction. W1094 AJR:197, December 2011

3 PROPELLER Technique in MRI of Shoulder and PROPELLER MRI. Interobserver agreement was assessed with Spearman rank correlation coefficient. Correlation coefficient values were interpreted according the charac terization of kappa statistics by Landis and Koch [8]. Values of indicated very good agreement, indicated good agreement, indicated moderate agreement, indicated fair agreement, and less than 0.20 indicated poor agreement. A software package (SPSS, version , SPSS) was used for all statistical calculations. Results Motion artifacts were significantly reduced in PROPELLER MRI compared with the standard TSE technique (Table 2 and Figs. 1 3) in the coronal oblique and sagittal oblique planes (median rating for PROPELLER, 5 for both observers [mean, ] vs median rating for standard, 4 5 for both observers [mean TSE, ]); the corresponding range of the Spearman rank correlation coefficient for both observers was Observer 1 rated motion artifacts with diagnostic impairment or a rating score of grade 3 or lower, respectively, in TABLE 2: Artifacts and Overall Image Quality Type of Artifact, Observer one patient (observer 2, 0 patients) on coronal oblique PROPELLER images compared with motion artifacts of grade 3 or lower in 33 patients (observer 2, 14 patients) on standard coronal oblique MRI (Table 3). Observer 1 noted motion artifacts grade 3 or lower in one patient (observer 2, 0 patients) on sagittal oblique PROPELLER images compared with 19 patients (observer 2, 21 patients) on standard sagittal oblique MRI. Magnetic susceptibility artifacts were significantly more pronounced in the PRO- PELLER group compared with the standard TSE group in the coronal oblique plane for observer 1 (Fig. 4) and in the sagittal plane for both observers (median PROPELLER, 5 [mean, ] vs median standard, 5 [mean TSE, ]). For reader 2, susceptibility artifacts were more pronounced in the PRO- PELLER group in the coronal oblique plane without statistically significance. The range of the Spearman rank correlation coefficient of susceptibility artifacts for both observer was Observer 1 noted susceptibility artifacts with diagnostic impairment or a rating score of grade 3 or lower, respectively, Coronal Oblique View, Technique in 12 patients (observer 2, nine patients) on sagittal oblique PROPELLER images compared with eight patients (observer 2, three patients) on standard sagittal oblique MRI. Pulsation artifacts were significantly more pronounced on standard TSE MRI compared with the PROPELLER technique in the coronal oblique plane for both observers and in the sagittal oblique plane for observer 2 (median PROPELLER, 5 [mean, ] vs median standard TSE, 4 5 [mean TSE, ]), whereas there was no significant difference for both evaluated techniques in the sagittal oblique plane for observer 1. The range of the Spearman rank correlation coefficient of pulsation artifacts for both observers was between 0.16 and Observer 1 rated pulsation artifacts with diagnostic impairment or a rating score of grade 3 or lower, respectively, in 0 patients (observer 2, 0 patients) on coronal oblique PROPELLER images compared with motion artifacts in one patient (observer 2, 13 patients) on standard coronal oblique MRI. The overall image quality was significantly better for the sequences with PROPELLER compared with the standard technique for Sagittal Oblique Standard PROPELLER p Standard PROPELLER p Motion artifacts (n = 111) Observer 1 4 (4.10) 5 (4.87) < (4.23) 5 (4.95) < Observer 2 5 (4.50) 5 (4.97) < (4.23) 5 (4.98) < Spearman correlation coefficient 0.59 a 0.45 a 0.49 a 0.33 a Susceptibility artifacts (n = 111) Observer 1 5 (4.68) 5 (4.62) (4.77) 5 (4.68) Observer 2 5 (4.83) 5 (4.80) (4.87) 5 (4.78) Spearman correlation coefficient 0.76 a 0.79 a 0.72 a 0.75 a Pulsation artifacts (n = 111) Observer 1 5 (4.92) 5 (5.00) (4.99) 5 (4.98) 0.32 Observer 2 4 (3.99) 5 (4.94) < (3.99) 5 (4.97) < Spearman correlation coefficient 0.15 b Not possible 0.16 c 0.02 d Overall image quality (n = 111) Observer 1 4 (3.83) 5 (4.57) < (4.06) 5 (4.68) < Observer 2 4 (3.85) 5 (4.86) < (3.60) 5 (4.87) < Spearman correlation coefficient 0.43 a 0.52 a 0.40 a 0.44 a Note Except where noted otherwise, data are median (mean) ratings by the observers. p values were calculated with Wilcoxon signed rank test. PROPELLER = periodically rotated overlapping parallel lines with enhanced reconstruction. a p < b p = c p = d p = AJR:197, December 2011 W1095

4 Dietrich et al. both observers and for both planes (median PROPELLER, 5 [mean, ] vs median standard TSE, 4 [mean TSE, ]). The corresponding range of the Spearman rank correlation coefficient for both observers was Observer 1 noted an overall image quality with diagnostic impairment or a rating score of grade 3 or lower, respectively, in nine patients (observer 2, two patients) on sagittal oblique PROPELLER images compared with 23 patients (observer 2, 40 patients) on standard sagittal oblique MRI. Delineation of all addressed anatomic structures (Table 4) was significantly better for the sequences with PROPELLER compared with the standard technique for both observers and for both planes (median PROPELLER, 5 vs median standard TSE, 4), and the corresponding range of the Spearman rank correlation coefficient for both observer was Interobserver agreement was good for the rating of susceptibility artifacts on standard and PROPELLER MRI in both imaging planes and for the rating of the articular and bursal supraspinatus tendon on PROPELLER MRI. Interobserver agreement was moderate for the rating of motion artifacts on standard MRI in both planes and coronal oblique PROPELLER MRI, for the rating of overall image quality on PROPELLER MRI in both A Fig year-old man. A, Coronal oblique intermediate-weighted turbo spin-echo image with fat suppression using standard technique shows minor motion artifacts and minor diagnostic impairment. B, With periodically rotated overlapping parallel lines with enhanced reconstruction, no motion artifacts are present. planes and coronal oblique standard MRI, for the rating of articular supraspinatus tendon on standard MRI, and for the rating of biceps anchor on standard and PROPELLER images. The interobserver agreement was fair and poor for the remaining available ratings. TABLE 3: Artifacts and Overall Image Quality Type of Artifact, Observer Motion artifacts (n = 111) Coronal Oblique View, Technique Sagittal Oblique Standard PROPELLER Standard PROPELLER Observer Observer Susceptibility artifacts (n = 111) Observer Observer Pulsation artifacts (n = 111) Observer Observer Overall image quality (n = 111) Discussion With the use of PROPELLER for MRI of the shoulder, motion artifacts were significantly reduced. The overall image quality and the delineation for all investigated anatomic structures could be improved compared with Observer Observer Note Data are number of patients with a score of grade 3 or lower. PROPELLER = periodically rotated overlapping parallel lines with enhanced reconstruction. B W1096 AJR:197, December 2011

5 PROPELLER Technique in MRI of Shoulder TABLE 4: Results for Delineation of Anatomic Structures View, Anatomic Structure, Observer the standard technique. PROPELLER minimized motion artifacts in most cases, with important influence on the image quality and the diagnostic value of MRI. The number of sequences with diagnostic impairment or a rating score of grade 3 or lower, respectively, for the overall image quality in the coronal oblique plane of observer 1 was reduced from 44 for the standard sequence to 11 for the PROPELLER technique. This indicates that the PROPELLER technique is an important technique in patients with substantial motion artifacts. The PROPELLER technique was first described to correct motion artifacts in imaging of the brain and to correct respiratory motion in free-breathing cardiac MRI by Pipe in 1999 [1]. In that article, Pipe stated that the PROPELLER technique reduces motion artifacts without the use of a dedicated motion correction. Later studies confirmed a significant reduction of motion artifacts caused by involuntary patient movement with the PROPELLER technique compared with standard MRI sequences in the chest, abdomen, spine, and brain in pediatric patients and those with stroke [1 7]. Pipe stated that the PROPELLER technique is well suited for imaging of moving objects because of its inherent ability to reject some of the in-plane as well as through-plane motion and its inherent averaging of the remaining data inconsistencies [1]. This is explained by data oversampling at the center region of k-space, which requires an additional factor of π/2 imaging time over conventional MRI techniques [1]. The disadvantage of the longer acquisition time of the PROPELLER technique varied from 28 seconds up to 1 minute 18 seconds for each sequence in our study. In daily practice, this might partially be compensated, because repetition of sequences may become unnecessary. Furthermore, redundant data sampling of PROPELLER MRI also increases the signal-to-noise ratio [1]. Fellner et al. [7] evaluated sagittal T2-weighted PROPELLER MRI in the cervical spine and found significantly increased signal-to-noise ratio in the vertebral body and spinal cord for the PROPELLER technique compared with the standard TSE sequence; however, there was no significant difference in the signalto-noise ratio in the vertebral disk and CSF between both techniques in the quantitative signal-to-noise ratio evaluation. Metallic implants, such as anchors used for rotator cuff repair, reduce the MRI quality because of susceptibility artifacts. These susceptibility artifacts were larger with the PROPELLER technique than with the standard TSE technique in our study. In contrast to standard TSE sequences, the direction of the phase and frequency readout gradients in PROPELLER sequences are permanently rotating in parallel lines in a 360 pattern around the center of the k-space; therefore, the phase and frequency readout gradients cannot be influenced by the operator of the MRI scanner. Standard Technique PROPELLER Coronal oblique Supraspinatus tendon articular (n = 104) Observer 1 4 (3.97) 5 (4.48) < Observer 2 4 (3.72) 5 (4.70) < Spearman correlation coefficient 0.49 a 0.65 a Supraspinatus tendon bursal (n = 104) Observer 1 4 (4.19) 5 (4.57) < Observer 2 4 (3.83) 5 (4.79) < Spearman correlation coefficient 0.32 b 0.62 a Biceps anchor (n = 95) Observer 1 4 (4.00) 5 (4.46) < Observer 2 4 (3.78) 5 (4.85) < Spearman correlation coefficient 0.51 a 0.41 a Glenohumeral cartilage (n = 111) Observer 1 4 (3.81) 5 (4.48) < Observer 2 4 (3.84) 5 (4.92) < Spearman correlation coefficient 0.37 a 0.26 c Sagittal oblique Biceps tendon in the rotator cuff interval (n = 95) Observer 1 4 (4.16) 5 (4.74) < Observer 2 4 (3.86) 5 (4.71) < Spearman correlation coefficient 0.22 d 0.14 e Acromioclavicular joint (n = 111) Observer 1 4 (4.38) 5 (4.76) < Observer 2 4 (3.82) 5 (4.85) < Spearman correlation coefficient 0.37 a 0.10 f Note Except where noted otherwise, data are median (mean) ratings by the observers. p values were calculated with Wilcoxon signed rank test. PROPELLER = periodically rotated overlapping parallel lines with enhanced reconstruction. a p < b p = c p = d p = e p = f p = However, in a recent study assessing PROPELLER with sagittal T2-weighted MRI of the cervical spine, no significant differences were seen for metallic artifacts compared with the standard technique [7]. These missing advantages of PROPELLER MRI compared with standard techniques regarding susceptibility artifacts in our study and in the cervical spine are in contrast to study findings of reduced susceptibility artifacts with echo-planar diffusion-weighted imaging using PROPELLER [9 15]. These p AJR:197, December 2011 W1097

6 Dietrich et al. different results might be explained by the investigated TSE standard sequences in our study, which are more robust for susceptibility artifacts than echo-planar diffusionweighted imaging. Pulsation artifacts from the axillary artery and vein were a minor source of artifacts in our study, causing no more than mild and slightly disturbing artifacts; however, these minor artifacts were reduced significantly with the PROPELLER technique for observer 1 in the coronal plane and for observer 2 in both planes. Michaely et al. [16] reported no disturbing pulsation artifacts of major vessels in A renal MRI using the PROPELLER technique in contrast to standard TSE. The authors concluded that the lower overall image quality rating of the standard TSE group was mainly caused by the increased number of disturbing motion and pulsation artifacts with the standard sequence. Similarly, pulsation artifacts as Fig year-old woman. A, Sagittal oblique T2-weighted turbo spin-echo sequence with fat suppression with standard MRI technique shows moderate motion artifacts. B, With periodically rotated overlapping parallel lines with enhanced reconstruction, motion artifacts are reduced. B A Fig year-old woman with full-thickness tear of supraspinatus tendon and tendon retraction to level of glenoid. A, On standard coronal oblique intermediate-weighted turbo spin-echo sequence with fat suppression, severe motion artifacts are present. Image is nondiagnostic. B, With periodically rotated overlapping parallel lines with enhanced reconstruction technique, motion artifacts are reduced. Image has diagnostic quality. B W1098 AJR:197, December 2011

7 PROPELLER Technique in MRI of Shoulder a result of blood flow in brain imaging were significantly reduced using PROPELLER MRI [17]. Flow phenomena of the CSF in the cervical spine were significantly reduced with the PROPLLER technique compared with the standard technique [7, 18]. These flow phenomena cause hypointense areas within the CSF, which sometimes can cause diagnostic difficulties. Attenberger et al. [19] stated that PROPELLER is not compatible with T1- weighted spin-echo sequences. PROPELLER implemented on T1-weighted MRI sequences results in low image quality because of the inaccurate estimation of motion [20]. Therefore, recently published PROPELLER sequences were mainly T2-weighted sequences, diffusion-weighted sequences, T2-weighted FLAIR, and T1-weighted FLAIR sequences [3, 7, 9, 11, 16 18, 21]. Our study has limitations. We assessed image artifacts and image quality, but we did not compare the diagnostic accuracy of the standard and the PROPELLER sequence. The effect of PROPELLER on T1-weighted images was not tested in this study. As a consequence of the favorable results of this study, we introduced PROPELLER A Fig year-old man with four metallic anchors in humeral head after subscapularis tendon repair. A and B, Coronal oblique intermediate-weighted turbo spin-echo images with fat suppression without (A) and with (B) periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER). Susceptibility artifacts are seen with both techniques. Susceptibility artifacts are slightly more pronounced with PROPELLER (arrows, B). sequences into our routine shoulder MRI protocol as coronal oblique intermediateweighted fat-saturated and sagittal oblique T2-weighted fat-saturated MRI instead the previously used standard sequences on our 1.5- and 3-T MRI scanners. We conclude that the PROPELLER technique for MRI of the shoulder reduces the number of sequences with diagnostic impairment because of motion artifacts and increases image quality compared with standard TSE sequences. Therefore, the PROPELLER technique for MRI of the shoulder is especially beneficial for patients with a higher likelihood for motion artifacts, such as uncooperative and claustrophobic patients. On the other hand, postoperative susceptibility artifacts associated with metallic implants are more pronounced with PROPELLER than in standard TSE MRI. The acquisition time increases compared with standard TSE sequences. References 1. 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, Karis JP, Farthing V, Heiserman JE. Brain imaging in the unsedated pediatric patient: comparison of periodically rotated overlapping parallel lines with enhanced reconstruction and single-shot fast spin-echo sequences. AJNR 2003; 24: Alibek S, Adamietz B, Cavallaro A, et al. Contrast-enhanced T1-weighted fluid-attenuated inversion-recovery BLADE magnetic resonance imaging of the brain: an alternative to spin-echo technique for detection of brain lesions in the unsedated pediatric patient? Acad Radiol 2008; 15: Hirokawa Y, Isoda H, Maetani Y, Arizono S, Shimada K, Togashi K. MRI artifact reduction and quality improvement in the upper abdomen with PROPELLER and prospective acquisition correction (PACE) technique. AJR 2008; 191: Hirokawa Y, Isoda H, Maetani Y, 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: Tamhane AA, Arfanakis K. Motion correction in periodically-rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) and tur- B AJR:197, December 2011 W1099

8 Dietrich et al. boprop MRI. Magn Reson Med 2009; 62: tion to high-resolution diffusion tensor imaging. enhanced MR imaging of the brain using T1-7. Fellner C, Menzel C, Fellner F, et al. BLADE in Magn Reson Med 2006; 56: weighted FLAIR with BLADE compared with a sagittal T2-weighted MR imaging of the cervical 13. Lehmann P, Saliou G, Brochart C, et al. 3T MR conventional spin-echo sequence. Eur Radiol 2008; spine. AJNR 2010; 31: imaging of postoperative recurrent middle ear 18: Landis JR, Koch GG. An application of hierar- cholesteatomas: value of periodically rotated over- 18. Ragoschke-Schumm A, Schmidt P, Schumm J, et chical kappa-type statistics in the assessment of majority agreement among multiple observers. Biometrics 1977; 33: 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: Wang FN, Huang TY, Lin FH, et al. PROPELLER EPI: an MRI technique suitable for diffusion tensor imaging at high field strength with reduced geometric distortions. Magn Reson Med 2005; 54: Adachi M, Kabasawa H, Kawaguchi E. Depiction of the cranial nerves within the brain stem with use of PROPELLER multishot diffusion-weighted imaging. AJNR 2008; 29: Chuang TC, Huang TY, Lin FH, et al. PROPEL- LER-EPI with parallel imaging using a circularly symmetric phased-array RF coil at 3.0 T: applica- lapping parallel lines with enhanced reconstruction diffusion-weighted MR imaging. AJNR 2009; 30: Kiryu S, Watanabe M, Kabasawa H, Akahane M, Aoki S, Ohtomo K. Evaluation of super paramagnetic iron oxide-enhanced diffusionweighted PROPELLER T2-fast spin echo magnetic resonance imaging: preliminary experience. J Comput Assist Tomogr 2006; 30: Deng J, Miller F, Salem R, Omary R, Larson A. Multishot diffusion-weighted PROPELLER magnetic resonance imaging of the abdomen. Invest Radiol 2006; 41: 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: Naganawa S, Satake H, Iwano S, et al. Contrast- al. Decreased CSF-flow artefacts in T2 imaging of the cervical spine with periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER/BLADE). Neuroradiology 2011; 53: 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: Feng Y, Chen W. New algorithm for extracting motion information from PROPELLER data and head motion correction in T1-weighted MRI. Conf Proc IEEE Eng Med Biol Soc 2005; 2: Alkan O, Kizilkiliç 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; 15:75 80 W1100 AJR:197, December 2011