- Email: [email protected]
T2-weighted MRI of the Upper Abdomen:
T2-weighted MRI of the Upper Abdomen: Comparison of Four Fat-Suppressed T2-weighted Sequences Including PROPELLER (BLADE) Technique Sibel Bayramoglu, MD, O¨zgu¨r Kilickesmez, MD, Tan Cimilli, MD, Arda Kayhan, MD, Gu¨lseren Yirik, MD, Filiz Islim, MD, Sedat Alibek, MD Rationale and Objectives: The aim of this study was to compare four different fat-suppressed T2-weighted sequences with different techniques with regard to image quality and lesion detection in upper abdominal magnetic resonance imaging (MRI) scans. Materials and Methods: Thirty-two consecutive patients referred for upper abdominal MRI for the evaluation of various suspected pathologies were included in this study. Different T2-weighted sequences (free-breathing navigator-triggered turbo spin-echo [TSE], freebreathing navigator-triggered TSE with restore pulse (RP), breath-hold TSE with RP, and free-breathing navigator-triggered TSE with RP using the periodically rotated overlapping parallel lines with enhanced reconstruction technique [using BLADE, a Siemens implementation of this technique]) were used on all patients. All images were assessed independently by two radiologists. Assessments of motion artifacts; the edge sharpness of the liver, pancreas, and intrahepatic vessels; depictions of the intrahepatic vessels; and overall image quality were performed qualitatively. Quantitative analysis was performed by calculation of the signal-to-noise ratios for liver tissue and gallbladder as well as contrast-to-noise ratios of liver to spleen. Results: Liver and gallbladder signal-to-noise ratios as well as liver to spleen contrast-to-noise ratios were significantly higher (P < .05) for the BLADE technique compared to all other sequences. In qualitative analysis, the severity of motion artifacts was significantly lower with T2-weighted free-breathing navigator-triggered BLADE sequences compared to other sequences (P < .01). The edge sharpness of the liver, pancreas, and intrahepatic vessels; depictions of the intrahepatic vessels; and overall image quality were significantly better with the BLADE sequence (P < .05). Conclusion: The T2-weighted free-breathing navigator-triggered TSE sequence with the BLADE technique is a promising approach for reducing motion artifacts and improving image quality in upper abdominal MRI scans. Key Words: MRI; fat-suppression; BLADE; liver. ªAUR, 2010
F
or liver magnetic resonance imaging (MRI), highquality T2-weighted (T2W) images are necessary for the detection and characterization of focal hepatic lesions. Various MRI techniques have been investigated to achieve optimal image quality (1–8). Recent technical developments such as powerful gradient systems and receiver coils with higher sensitivities have significantly enhanced image quality. However, motion artifacts remain a central problem of upper abdominal MRI scans, being the main cause of image quality degradation, and great efforts have been made to achieve high-quality T2W images without motion artifacts. Today, T2W turbo spin-echo (TSE) sequences are generally
Acad Radiol 2010; 17:368–374 From the Department of Radiology, Bakırkoy Dr Sadi Konuk Research and Training Hospital, Istanbul, Turkey (S.B., O.K., T.C., A.K., G.Y., F.I.); and the Radiology Institute, University of Erlangen-Nuremberg, Maximiliansplatz 1, 91054 Erlangen, Germany (S.A.). Received September 2, 2009; accepted October 1, 2009. Address correspondence to: S.A. e-mail: sedat.alibek@ uk-erlangen.de ªAUR, 2010 doi:10.1016/j.acra.2009.10.015
368
considered to be the standard imaging sequence for the evaluation of liver parenchyma. Breath-hold (BH) imaging with rapid imaging techniques has been a useful method that can reduce respiratory motion artifacts, providing shorter acquisition times and thereby increasing patient compliance and throughput. TSE and half-Fourier acquisition single-shot TSE are widely used methods of BH T2W imaging that permit scanning of the entire liver in a single or a few BH steps (9–11). Another approach for reducing respiratory motion artifacts is to use triggering techniques that synchronize anatomic data acquisition with the respiratory cycle using a belt placed on the patient’s abdomen (12,13). Respiratory triggering techniques provide high tissue contrast and allow images to be obtained at thin slice thicknesses. Recently, a new free-breathing, navigator-triggered T2W imaging technique, called Prospective Acquisition Correction (PACE; Siemens Medical Systems, Erlangen, Germany), has been developed. In this technique, navigator echoes are used to monitor respiratory motion instead of a respiratory belt. The patient does not need to be particularly prepared,
Academic Radiology, Vol 17, No 3, March 2010
T2W MRI OF THE UPPER ABDOMEN
TABLE 1. Summary of Imaging Parameters for Each Fat-suppressed T2W Sequence
Parameter Repetition time (ms) Echo time (ms) Matrix (frequency phase) Field of view (%) Bandwidth (kHz) Turbo factor Flip angle ( ) Number of signals acquired Number of sections acquired
T2W-PACE*-TSE with RP
T2W-PACE-TSE
BH-T2W-TSE with RP
T2W-BLADE* with PACE and RP
4600 90 320 153 350 260–300 260 19 150 2 35
4600 90 320 153 350 260–300 260 19 150 2 35
3400 90 256 129 350 260–300 260 27 150 1 24
4000 100 256 350 350 362 35 150 1 35
BH, breath-hold; PACE, Prospective Acquisition Correction; RP, restore pulse; TSE, turbo spin-echo; T2W, T2-weighted. *Siemens Medical Systems (Erlangen, Germany).
and interruptions of the examination due to dislocation of the monitoring device do not occur. A few recent studies have shown the usefulness of this technique for reducing respiratory motion artifacts in T2W imaging of the upper abdomen (14–16). A limitation of the BH T2W TSE sequence is a loss in softtissue resolution and decreased signal-to-noise ratio (SNR) because of proton saturation with short repetition times. A modified TSE technique using an additional restore pulse (RP) was reported to provide excellent T2W images with higher lesion-to-liver contrast and fewer image artifacts than the respiratory-triggered fast spin-echo and single-shot fast spin-echo techniques, respectively (17–19). Conversely, a recent study reported the accuracy of navigator-triggered TSE and respiratory-triggered TSE sequences to be superior to that of BH TSE with RP and half-Fourier acquisition single-shot TSE for the detection of all focal solid hepatic lesions (15). Although BH TSE with RP has been evaluated, this sequence has never been used with a navigator-triggered method. BLADE, a Siemens implementation of the periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) technique, was applied to reduce motion artifacts for brain imaging and has been reported to improve image quality (20–24). Recently, the BLADE technique was also used for examinations of the upper abdomen (24,25), but it has never been used with RP to our knowledge. In this study, we aimed to compare four different T2W TSE sequences, including BLADE, for upper abdominal imaging, with a focus on image quality. MATERIALS AND METHODS Patients
From November 2007 to April 2008, 35 consecutive adult patients referred for MRI of the upper abdomen for the evaluation of various suspected pathologies were included in this study and examined. Three patients were retrospectively excluded from the study because of lack of cooperation. Therefore, a total of 32 patients (20 women, 16 men; age range, 29–71
years; mean age, 50 16 years) were included. Our institutional review board approved this study, and written informed consent was obtained from all patients prior to MRI scans. MRI Protocol
MRI was performed on a 1.5-T scanner (Magnetom Avanto; Siemens Medical Systems) with a 33 mT/m maximum gradient capability using a dedicated six-element phased-array body coil (12 elements with spine coil). Free-breathing navigator-triggered T2W TSE (T2W-PACE-TSE), freebreathing navigator-triggered TSE with RP (T2W-PACETSE with RP), BH TSE with RP (BH-T2W-TSE with RP), and free-breathing navigator-triggered T2W BLADE with RP (T2W-BLADE with RP) sequences were acquired in all patients. All T2W images were obtained in the axial plane with a section thickness of 6 mm and a 10% to 15% intersection gap. The scan range covered the entire liver in all patients. The fat-suppression technique was used for all four T2W imaging sequences. Integrated parallel acquisition techniques were not used. The technical parameters of the sequences are listed in Table 1. For BH-T2W-TSE with RP, images were acquired within single or multiple BHs, with each BH period lasting <20 seconds. The T2W-PACE-TSE with RP sequence is a two-dimensional, navigator-based technique performed in real time during data acquisition using a navigator placed on the dome of the right hemidiaphragm. The respiratory trace sampled with the navigator portion of the sequence is used to synchronize the data acquisition with the patient’s respiratory cycle. Repetition times ranged from 3000 to 4600 ms, depending on the respiratory intervals. This cycle is repeated until all anatomic data have been acquired. The PROPELLER method (BLADE) acquires a certain number of so-called blades that rotate around the center of the k-space. Each blade consists of the smallest number of phase-encoding lines of a conventional rectilinear k-space trajectory that are filled with the multiple echo-train acquisition after a single radiofrequency signal acquisition. Each 369
BAYRAMOGLU ET AL
Academic Radiology, Vol 17, No 3, March 2010
TABLE 2. Results of Image Data Analysis for Each Fat-suppressed T2W Sequence
Sequence Reader 1 T2W-PACEy-TSE with RP T2W-PACE-TSE BH-T2W-TSE with RP T2W-BLADEy with PACE and RP P Reader 2 T2W-PACE-TSE with RP T2W-PACE-TSE BH-T2W-TSE with RP T2W-BLADE with PACE and RP P
Motion Artifacts
Edge Sharpness of Liver, Pancreas, and Intrahepatic Vessels
Depiction of Intrahepatic Vessels
Overall Image Quality
3.88 0.52 3.74 0.48 3.24 0.43 4.84 0.47 *P2–P6 < .01 P1 > .05
4.12 0.43 4.00 0.50 3.40 0.50 4.92 0.40 *P2–P6 < .01 P1 > .05
4.20 0.50 4.00 0.64 3.52 0.50 4.88 0.33 *P1–P6 < .01
4.20 0.40 4.04 0.35 3.56 0.50 4.96 0.20 *P2–P6 < .01 P1 > .05
3.80 0.40 3.68 0.55 3.32 0.47 4.92 0.27 *P2–P6 < .01 P1 > .05
4.04 0.53 3.92 0.57 3.34 0.47 4.84 0.47 *P2–P6 < .01 P1 > .05
4.28 0.45 4.04 0.61 3.60 0.50 4.84 0.37 *P1–P6 < .01
4.24 0.43 4.04 0.45 3.52 0.50 4.88 0.43 *P1–P6 < .01
BH, breath-hold; P1, P value between first and second sequences; P2, P value between first and fourth sequences; P3, P value between first and third sequences; P4, P value between second and fourth sequences; P5, P value between second and third sequences; P6, P value between third and fourth sequences; PACE, Prospective Acquisition Correction; RP, restore pulse; TSE, turbo spin-echo; T2W, T2-weighted. Data are expressed as mean standard deviation. *P < .01. y Siemens Medical Systems (Erlangen, Germany).
blade was evenly rotated 11 times to cover the k-space completely. The width of each blade, which defines the number of phase-encoding lines per blade, was 35, and k-space coverage was 100%. The mean image acquisition times were 205 52 seconds for T2W-PACE-TSE with RP and T2W-PACE-TSE, 245 65 seconds for T2W-BLADE with PACE and RP, and 40 seconds for BH-T2W-TSE with RP. Image Analysis
In each subject, T2W-PACE-TSE, T2W-PACE-TSE with RP, BH-T2W-TSE with RP, and T2W-BLADE with RP images were acquired and evaluated qualitatively in terms of overall image quality, motion artifacts (ie, respiratory ghosting, vascular pulsation, peristalsis, and susceptibility), the depiction of the intrahepatic vessels, and the edge sharpness of the liver, pancreas, and intrahepatic vessels. In quantitative analysis, the mean signal intensities (SIs) of the liver, spleen, and gallbladder as well as background noise were measured, and the liver SNR (SIliver/SDnoise), liver-to-spleen contrast-to-noise ratio (CNR = [SIspleen SIliver]/SDnoise), and gallbladder SNR (SIgallbladder/SDnoise) were calculated, where SD is the standard deviation. All evaluated items were compared to one another statistically. Qualitative Evaluation
Assessments of images from all four sequences were performed by two independent readers with 4 and 5 years of experience in abdominal MRI. Both reviewers were unaware of the sequence parameters and patient information. The reviewers 370
graded the overall image quality of each T2W sequence using the following subjective five-point scale: 1 = unacceptable, 2 = poor, 3 = fair, 4 = good, and 5 = excellent. The presence of motion artifacts such as respiratory ghosting, vascular pulsation, peristalsis, and susceptibility was graded as 1 = severe, 2 = moderate, 3 = mild, 4 = minimal, or 5 = absent. The depiction of the intrahepatic vessels was assessed as 1 = unacceptable (invisible main portal vein), 2 = poor (only main portal vein visible), 3 = fair (only main branch of portal vein visible), 4 = good (some peripheral portal veins visible), or 5 = excellent (peripheral portal veins clearly visible). The edge sharpness of the liver, pancreas, and intrahepatic vessels was graded as 1 = unacceptable, 2 = poor, 3 = fair, 4 = good, or 5 = excellent. Quantitative Evaluation
Quantitative analysis was performed on the T2W magnetic resonance images with operator-defined region-of-interest (ROI) measurements of the mean SIs in the liver, spleen, and gallbladder as well as background noise. The SIs of the liver and spleen were measured in areas devoid of large vessels and prominent artifacts. The mean SIs of liver and spleen were determined with standard ROI measurements in three different sections from each set of images (the upper, middle, and lower parts of the spleen and liver). For all measurements, the size of the ROI was identical for all acquisition methods. The SD of background noise was measured in the largest possible ROI positioned in the phase-encoding direction outside the abdominal wall to account for any motion artifacts. The liver SNR, liver-to-spleen CNR, and gallbladder SNR were then calculated.
Academic Radiology, Vol 17, No 3, March 2010
T2W MRI OF THE UPPER ABDOMEN
Figure 1. Sample images of a patient with two hemangiomas in the liver: (a) T2-weighted (T2W) turbo spin-echo (TSE) Prospective Acquisition Correction (PACE); (b) T2W TSE PACE with restore pulse (RP); (c) T2W TSE breath-hold with RP; and (d) T2W BLADE PACE with RP.
Statistical Analysis
Statistical analysis was performed using NCSS 2007 and PASS 2008 statistical software (NCSS, Kaysville, UT). Friedman repeated-measures analysis of variance on ranks was used on the basis of negative results on tests assessing distribution normality. Wilcoxon’s signed-rank test was applied as a post hoc test for differences in multiple comparisons. In these analysis, P values < .05 were considered to indicate statistically significant differences. Reader agreement was assessed using Cohen’s k test. A value of 0 to 0.40 implied fair, 0.41 to 0.60 moderate, 0.61 to 0.80 substantial, and 0.81 to 1 almost perfect agreement. RESULTS Qualitative Analysis
Table 2 shows the results for motion artifacts; the edge sharpness of the liver, pancreas, and intrahepatic vessels; depictions of the intrahepatic vessels; and overall image quality as evaluated by the two observers for the four different sequences (Fig 1). The severity of motion artifacts induced by respiratory ghosting, vascular pulsation, peristalsis, and susceptibility was significantly lower in BLADE T2W TSE with RP images compared to the other sequences (P < .01). The edge sharpness of the liver, pancreas, and intrahepatic vessels and depictions of the intrahepatic vessels were also significantly better with this sequence (P < .05). There was no statistically significant difference between the T2W-PACE-TSE with RP and T2W-PACE-TSE sequences in terms of motion artifacts and the edge sharpness of the liver and pancreas. All examination criteria were significantly lower in the BH-T2W-TSE with RP sequence. Interobserver agreement for all evaluated items was judged almost perfect (Table 3).
TABLE 3. Results of Cohen’s k Statistical Analysis of Motion Artifacts, Edge Sharpness of the Liver and Pancreas, Depiction and Sharpness of Intrahepatic Vessels, and Overall Image Quality Between Readers Parameter Motion artifacts Edge sharpness of liver, pancreas, and intrahepatic vessels Depiction of intrahepatic vessels Overall image quality
k 88.2% 84.5% 86.6% 85.9%
Quantitative Analysis
The results of quantitative assessments of the mean background noise, liver SNR, liver-to-spleen CNR, and gallbladder SNR are summarized in Table 4. Differences among the sequences for these quantitative evaluation are shown in Table 5. The background noise with BH-T2W-TSE with RP was significantly lower than with the T2W-PACE-TSE with RP and BLADE sequences (P < .001). The mean liver SNR and liver-to-spleen CNR obtained with the BLADE sequence were significantly higher than those obtained using the other pulse sequences (P < .001). Similarly, the mean liver CNR was significantly higher with T2W-PACE-TSE with RP than with the T2W-PACE-TSE and BH-T2W-TSE with RP sequences. There was no statistically significant difference between the T2W-PACE-TSE and BH-T2W-TSE with RP sequences. The gallbladder SNR was highest with the BLADE sequence, but there was no statistically significant difference between the T2W-PACE-TSE with RP and BLADE. T2WPACE-TSE with RP was higher than the T2W-PACE-TSE and BH-T2W-TSE with RP sequences in terms of gallbladder 371
BAYRAMOGLU ET AL
Academic Radiology, Vol 17, No 3, March 2010
TABLE 4. Results of Quantitative Assessment of Background Noise, Liver SNR, Liver-to-Spleen CNR, and Gallbladder SNR Sequence T2W-PACE*-TSE with RP T2W-PACE-TSE T2W-BLADE* with PACE and RP BH-T2W-TSE with RP
Background Noise
Liver SNR
Liver-to-Spleen CNR
Gallbladder SNR
2.2 0.6 2.3 0.7 2.2 0.8 2.0 0.6
65.4 22.8 58.2 20.7 84.0 31.5 56.2 26.8
72.4 34.5 65.1 24.4 84.0 35.1 63.7 36.4
372.9 137.9 317.4 109.5 420.2 131.7 299.8 136.7
BH, breath-hold; CNR, contrast-to-noise ratio; PACE, Prospective Acquisition Correction; RP, restore pulse; SNR, signal-to-noise ratio; TSE, turbo spin-echo; T2W, T2-weighted. Data are expressed as mean standard deviation. *Siemens Medical Systems (Erlangen, Germany).
TABLE 5. Differences Between Sequences in Terms of Background Noise, Liver SNR, Liver-to-Spleen CNR, and Gallbladder SNR
Comparison T2W-BLADE* with PACE* and RP vs T2W-PACE-TSE with RP T2W-BLADE with PACE and RP vs T2W-PACE-TSE T2W-BLADE with PACE and RP vs BH-T2W-TSE with RP T2W-PACE-TSE with RP vs T2W-PACE-TSE T2W-PACE-TSE with RP vs BH-T2W-TSE with RP T2W-PACE-TSE vs BH-T2W-TSE with RP
Background Noise (P)
Liver SNR (P)
Liver-to-Spleen CNR (P)
Gallbladder SNR (P)
<.001 <.001 <.001 <.001 <.05 .740
<.001 <.001 <.001 <.001 <.05 .74
<.001 <.001 <.001 .07 <.05 .45
.08 <.01 <.001 <.001 <.001 .22
BH, breath-hold; CNR, contrast-to-noise ratio; PACE, Prospective Acquisition Correction; RP, restore pulse; SNR, signal-to-noise ratio; TSE, turbo spin-echo; T2W, T2-weighted. *Siemens Medical Systems (Erlangen, Germany).
SNR. There was no significant difference between the T2WPACE-TSE and BH-T2W-TSE with RP sequences in all quantitative evaluations. The mean image acquisition time for BH-T2W-TSE with RP was significantly shorter than for the other sequences (P < .001). There was no significant difference in the mean image acquisition time among other sequences. DISCUSSION Patient motion is a significant problem in MRI of the upper abdomen, leading to a reduction in image quality and a loss of diagnostic information. Motion caused by respiration and cardiac and vascular pulsation occurring during data acquisition causes image artifacts, a loss of resolution, and a reduction in SNR (26,27). It also may reduce anatomic details and lead to lesions’ being obscured in images. The BLADE technique offers advantages over other methods used for patient motion correction, because it uses inherent data to correct two main motions, in-plane rotation and translation, which occur in supine patients undergoing axial abdominal imaging (28). This is made possible by acquiring data in a series of concentric blades, each of which rotates through the center of the k-space. This offers two main benefits over conventional data collection. First, the central region of the k-space is sampled multiple times, and this method yields a better SNR by oversampling data at the center of the k-space. Second, data within this central region can be compared between each blade. If motion has occurred between the acquisition of each blade, data can be transposed to its 372
estimated stationary position, prior to the final image reconstruction. Therefore, artifact suppression is improved. Consequently, the BLADE sequence is less sensitive to motion than conventional sequences. Using the BLADE technique, image quality could be improved, especially in unsedated patients, without a loss of diagnostic information (21). In our study, we observed this sequence to provide the same advantages in upper abdominal images. There were significant differences among BLADE and the other sequences in terms of motion artifacts and image quality. We especially observed the depictions of intrahepatic vessels and the edge sharpness of solid organs to be superior with the BLADE sequence. The reason for this is a more uniform distribution of echo times throughout the k-space, resulting in improved image contrast. Several studies reporting improved visual image quality, reduced motion artifacts, and increased detectability of lesions in the brain and upper abdomen with PROPELLER (BLADE) sequences have been published (20–25,29). In our study also, the highest SNR, liver-to-spleen CNR, and gallbladder SNR were obtained with the BLADE sequence. This result shows that oversampling data at the center of the k-space yields a better SNR with the BLADE technique. On BLADE images, a radial artifact occasionally appeared around a structure with high SI (22). Our experience with the BLADE technique suggests that these artifacts are not a significant problem, and the radial dispersion of motion artifacts may serve to further reduce visible motion. A limitation of the BH T2W TSE sequence is decreased SI due to proton saturation with short repetition times (30).
Academic Radiology, Vol 17, No 3, March 2010
However, a modified TSE sequence using an additional RP at the end of multiple refocusing pulses was reported to provide excellent T2W images with higher lesion-to-liver contrast than the TSE sequence (7,18,19). On the other hand, several authors have suggested that respiratory-triggered and navigator-triggered TSE are better than BH TSE with RP (15). Although BH TSE with RP has been evaluated, this sequence has never been used with a navigator-triggered method. In our study, liver and gallbladder SNRs obtained with T2WPACE-TSE with RP were significantly higher than those obtained using T2W-PACE-TSE. The benefit of an additional RP in the TSE sequence is higher T2 contrast. In our study, this feature was accentuated especially in fluid-containing structures, such as the gallbladder. There are also several potential advantages of navigator-triggered TSE over the BH technique. T2W-PACE-TSE with RP was also significantly better than BH-T2W-TSE with RP in terms of SNR, liver-to-spleen CNR, and gallbladder SNR. The higher spatial resolution in the magnetic resonance images obtained with the navigatorecho technique using a larger matrix size and two signal acquisitions may be a possible reason for the superior image technique. Image sharpness was also significantly better with this technique than with the BH technique. PACE is used to improve the image quality and high-resolution examinations with thinner slice thicknesses. This is especially useful for patients who have difficulty undergoing BH examinations or those under sedation. Our study had one limitation. Because only nine solid lesions were found in our series, we did not evaluate detectability, although the detection of solid tumors is an important indicator for image contrast in abdominal MRI. Our goal was the evaluation of image quality. Hirokawa et al (31) reported that T2W MRI with PACE and PROPELLER techniques and superparamagnetic iron oxide enhancement is a promising method with which to improve the detection of hepatic lesions. Further studies should be performed to consider lesion detection in larger series. In conclusion, our findings suggest that BLADE MRI offers reductions in motion artifacts and improvements in image quality over conventional sequences. Using the PACE method improved image quality and allowed high-resolution examination with thinner slice thicknesses. An additional RP provides better image quality because of the acquisition of higher contrast.
REFERENCES 1. Tang Y, Yamashita Y, Namimoto T, et al. Liver T2-weighted MR imaging: comparison of fast and conventional half-Fourier single-shot turbo spinecho, breath-hold turbo spin-echo, and respiratory-triggered turbo spinecho sequences. Radiology 1997; 203:766–772. 2. Ito K, Mitchell DG, Outwater EK, et al. Hepatic lesions: discrimination of nonsolid, benign lesions from solid, malignant lesions with heavily T2weighted fast spin-echo MR imaging. Radiology 1997; 204:729–737. 3. Matsuo M, Kanematsu M, Murakami T, et al. T-weighted MR imaging for focal hepatic lesion detection: supplementary value of breath-hold imaging with half-Fourier single-shot fast spin-echo and multishot spin-echo echoplanar sequences. J Magn Reson Imaging 2000; 12: 444–452.
T2W MRI OF THE UPPER ABDOMEN
4. Pawluk RS, Borrello JA, Brown JJ, et al. A prospective assessment of breath-hold fast spin echo and inversion recovery fast spin echo techniques for detection and characterization of focal hepatic lesions. Magn Reson Imaging 2000; 18:543–551. 5. Hori M, Murakami T, Kim T, et al. Single breath-hold T2-weighted MR imaging of the liver: value of single-shot fast spin-echo and multishot spinecho echoplanar imaging. AJR Am J Roentgenol 2000; 174:1423–1431. 6. Augui J, Vignaux O, Argaud C, et al. Liver: T2-weighted MR imaging with breath-hold fast–recovery optimized fast spin-echo compared with breath-hold half-Fourier and non-breath-hold respiratory-triggered fast spin-echo pulse sequences. Radiology 2002; 223:853–859. 7. Yoshikawa T, Mitchell DG, Hirota S, et al. Focal liver lesions: breathhold gradient and spin-echo T2-weighted imaging for detection and characterization. J Magn Reson Imaging 2006; 23:520–528. 8. Herborn CU, Vogt F, Lauenstein TC, et al. MRI of the liver: can true FISP replace HASTE? J Magn Reson Imaging 2003; 17:190–196. 9. Reinig JW. Breath-hold fast spin echo MR imaging of the liver: a technique for high-quality T2-weighted images. Radiology 1995; 194:303–304. 10. Gaa J, Hatabu H, Jenkins RL, et al. Liver masses: replacement of conventional T2-weighted spin-echo MR imaging with breath-hold MR imaging. Radiology 1996; 200:459–464. 11. Yu JS, Kim KW, Kim YH, et al. Comparison of multishot turbo spin echo and HASTE sequences for T2-weighted MRI of liver lesions. J Magn Reson Imaging 1998; 8:1079–1084. 12. Pauleit D, Textor J, Bachmann R, et al. Improving the detectability of focal liver lesions on T2-weighted MR images: ultrafast breath-hold or respiratory-triggered thin section MRI? J Magn Reson Imaging 2001; 14:128–133. 13. Kanematsu M, Hoshi H, Itoh K, et al. Focal hepatic lesion detection: comparison of four fat-suppressed T2-weighted MR imaging pulse sequences. Radiology 1999; 211:363–371. 14. Zech CJ, Herrmann KA, Huber A, et al. High-resolution MR-imaging of the liver with T2-weighted sequences using integrated parallel imaging: comparison of prospective motion correction and respiratory triggering. J Magn Reson Imaging 2004; 20:443–450. 15. Lee SS, Byun JH, Hong HS, et al. Image quality and focal lesion detection on T2-weighted MR imaging of the liver: comparison of two high-resolution free-breathing imaging techniques with two breath-hold imaging techniques. J Magn Reson Imaging 2007; 26:323–330. 16. Klessen C, Asbach P, Kroencke TJ, et al. Magnetic resonance imaging of the upper abdomen using a free-breathing T2-weighted turbo spin echo sequence with navigator triggered prospective acquisition correction. J Magn Reson Imaging 2005; 21:576–582. 17. Katayama M, Masui T, Kobayashi S, et al. Fat-suppressed T2-weighted MRI of the liver: comparison of respiratory-triggered fast spin-echo, breath-hold single-shot fast spin-echo, and breath-hold fast-recovery fast spin-echo sequences. J Magn Reson Imaging 2001; 14:439–449. 18. Huang J, Raman SS, Vuong N, et al. Utility of breath-hold fast-recovery fast spin-echo T2 versus respiratory-triggered fast spin-echo T2 in clinical hepatic imaging. AJR Am J Roentgenol 2005; 184:842–846. 19. Akın O, Schwartz LH, Webler A, et al. Evaluation of focal liver lesions: fastrecovery fast spin-echo T2-weighted MR imaging. Clin Imaging 2006; 30: 322–325. 20. Forbes KP, Pipe JG, Bird CR, et al. PROPELLER MRI: clinical testing of a novel technique for quantification and compensation of head motion. J Magn Reson Imaging 2001; 14:215–222. 21. Forbes KP, Pipe JG, Karis JP, et al. 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 Am J Neuroradiol 2003; 24:794–798. 22. Wintersperger BJ, Runge VM, Biswas J, et al. Brain magnetic resonance imaging at 3 Tesla using BLADE compared with standard rectilinear data sampling. Invest Radiol 2006; 41:586–592. 23. 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; 15:986–995. 24. Kiryu S, Watanabe M, Kabasawa H, et al. Evaluation of super paramagnetic iron oxide-enhanced diffusion-weighted PROPELLER T2-fast spin echo magnetic resonance imaging. J Comput Assist Tomogr 2006; 30: 197–200. 25. Hirokawa Y, Isoda H, Maetani YS, et al. MRI artifact reduction and quality improvement in the upper abdomen with PROPELLER and Prospective Acquisition Correction (PACE) technique. AJR Am J Roentgenol 2008; 191:1154–1158.
373
BAYRAMOGLU ET AL
26. Bellon EM, Haacke EM, Coleman PE, et al. MR artifacts: a review. AJR Am J Roentgenol 1987; 148:1251–1258. 27. Low RN, Alzate GD, Schimakawa A. Motion suppression in MR imaging of the liver: comparison of respiratory-triggered and non-triggered fast spinecho sequences. AJR Am J Roentgenol 1997; 168:225–231. 28. Pipe JG. Motion correction with PROPELLER MRI: application to head motion and free-breathing cardiac imaging. Magn Reson Med 1999; 42: 963–969. 29. Hirokawa Y, Isoda H, Maetani YS, et al. Evaluation of motion correction effect and image quality with the periodically rotated overlapping parallel
374
Academic Radiology, Vol 17, No 3, March 2010
lines with enhanced reconstruction (PROPELLER) (BLADE) and parallel imaging acquisition technique in the upper abdomen. J Magn Reson Imaging 2008; 28:957–962. 30. Yamashita Y, Yamamato H, Namimoto T, et al. Phased array breath-hold versus non-breath-hold MR imaging of focal liver lesions: a prospective comparative study. J Magn Reson Imaging 1997; 7:292–297. 31. Hirokawa Y, Isoda H, Maetani YS, et al. Hepatic lesions: improved image quality and detection with the periodically rotated overlapping parallel lines with enhanced reconstruction technique-evaluation of SPIO-enhanced T2-weighted MR images. Radiology 2009; 251:388–397.
F
or liver magnetic resonance imaging (MRI), highquality T2-weighted (T2W) images are necessary for the detection and characterization of focal hepatic lesions. Various MRI techniques have been investigated to achieve optimal image quality (1–8). Recent technical developments such as powerful gradient systems and receiver coils with higher sensitivities have significantly enhanced image quality. However, motion artifacts remain a central problem of upper abdominal MRI scans, being the main cause of image quality degradation, and great efforts have been made to achieve high-quality T2W images without motion artifacts. Today, T2W turbo spin-echo (TSE) sequences are generally
Acad Radiol 2010; 17:368–374 From the Department of Radiology, Bakırkoy Dr Sadi Konuk Research and Training Hospital, Istanbul, Turkey (S.B., O.K., T.C., A.K., G.Y., F.I.); and the Radiology Institute, University of Erlangen-Nuremberg, Maximiliansplatz 1, 91054 Erlangen, Germany (S.A.). Received September 2, 2009; accepted October 1, 2009. Address correspondence to: S.A. e-mail: sedat.alibek@ uk-erlangen.de ªAUR, 2010 doi:10.1016/j.acra.2009.10.015
368
considered to be the standard imaging sequence for the evaluation of liver parenchyma. Breath-hold (BH) imaging with rapid imaging techniques has been a useful method that can reduce respiratory motion artifacts, providing shorter acquisition times and thereby increasing patient compliance and throughput. TSE and half-Fourier acquisition single-shot TSE are widely used methods of BH T2W imaging that permit scanning of the entire liver in a single or a few BH steps (9–11). Another approach for reducing respiratory motion artifacts is to use triggering techniques that synchronize anatomic data acquisition with the respiratory cycle using a belt placed on the patient’s abdomen (12,13). Respiratory triggering techniques provide high tissue contrast and allow images to be obtained at thin slice thicknesses. Recently, a new free-breathing, navigator-triggered T2W imaging technique, called Prospective Acquisition Correction (PACE; Siemens Medical Systems, Erlangen, Germany), has been developed. In this technique, navigator echoes are used to monitor respiratory motion instead of a respiratory belt. The patient does not need to be particularly prepared,
Academic Radiology, Vol 17, No 3, March 2010
T2W MRI OF THE UPPER ABDOMEN
TABLE 1. Summary of Imaging Parameters for Each Fat-suppressed T2W Sequence
Parameter Repetition time (ms) Echo time (ms) Matrix (frequency phase) Field of view (%) Bandwidth (kHz) Turbo factor Flip angle ( ) Number of signals acquired Number of sections acquired
T2W-PACE*-TSE with RP
T2W-PACE-TSE
BH-T2W-TSE with RP
T2W-BLADE* with PACE and RP
4600 90 320 153 350 260–300 260 19 150 2 35
4600 90 320 153 350 260–300 260 19 150 2 35
3400 90 256 129 350 260–300 260 27 150 1 24
4000 100 256 350 350 362 35 150 1 35
BH, breath-hold; PACE, Prospective Acquisition Correction; RP, restore pulse; TSE, turbo spin-echo; T2W, T2-weighted. *Siemens Medical Systems (Erlangen, Germany).
and interruptions of the examination due to dislocation of the monitoring device do not occur. A few recent studies have shown the usefulness of this technique for reducing respiratory motion artifacts in T2W imaging of the upper abdomen (14–16). A limitation of the BH T2W TSE sequence is a loss in softtissue resolution and decreased signal-to-noise ratio (SNR) because of proton saturation with short repetition times. A modified TSE technique using an additional restore pulse (RP) was reported to provide excellent T2W images with higher lesion-to-liver contrast and fewer image artifacts than the respiratory-triggered fast spin-echo and single-shot fast spin-echo techniques, respectively (17–19). Conversely, a recent study reported the accuracy of navigator-triggered TSE and respiratory-triggered TSE sequences to be superior to that of BH TSE with RP and half-Fourier acquisition single-shot TSE for the detection of all focal solid hepatic lesions (15). Although BH TSE with RP has been evaluated, this sequence has never been used with a navigator-triggered method. BLADE, a Siemens implementation of the periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) technique, was applied to reduce motion artifacts for brain imaging and has been reported to improve image quality (20–24). Recently, the BLADE technique was also used for examinations of the upper abdomen (24,25), but it has never been used with RP to our knowledge. In this study, we aimed to compare four different T2W TSE sequences, including BLADE, for upper abdominal imaging, with a focus on image quality. MATERIALS AND METHODS Patients
From November 2007 to April 2008, 35 consecutive adult patients referred for MRI of the upper abdomen for the evaluation of various suspected pathologies were included in this study and examined. Three patients were retrospectively excluded from the study because of lack of cooperation. Therefore, a total of 32 patients (20 women, 16 men; age range, 29–71
years; mean age, 50 16 years) were included. Our institutional review board approved this study, and written informed consent was obtained from all patients prior to MRI scans. MRI Protocol
MRI was performed on a 1.5-T scanner (Magnetom Avanto; Siemens Medical Systems) with a 33 mT/m maximum gradient capability using a dedicated six-element phased-array body coil (12 elements with spine coil). Free-breathing navigator-triggered T2W TSE (T2W-PACE-TSE), freebreathing navigator-triggered TSE with RP (T2W-PACETSE with RP), BH TSE with RP (BH-T2W-TSE with RP), and free-breathing navigator-triggered T2W BLADE with RP (T2W-BLADE with RP) sequences were acquired in all patients. All T2W images were obtained in the axial plane with a section thickness of 6 mm and a 10% to 15% intersection gap. The scan range covered the entire liver in all patients. The fat-suppression technique was used for all four T2W imaging sequences. Integrated parallel acquisition techniques were not used. The technical parameters of the sequences are listed in Table 1. For BH-T2W-TSE with RP, images were acquired within single or multiple BHs, with each BH period lasting <20 seconds. The T2W-PACE-TSE with RP sequence is a two-dimensional, navigator-based technique performed in real time during data acquisition using a navigator placed on the dome of the right hemidiaphragm. The respiratory trace sampled with the navigator portion of the sequence is used to synchronize the data acquisition with the patient’s respiratory cycle. Repetition times ranged from 3000 to 4600 ms, depending on the respiratory intervals. This cycle is repeated until all anatomic data have been acquired. The PROPELLER method (BLADE) acquires a certain number of so-called blades that rotate around the center of the k-space. Each blade consists of the smallest number of phase-encoding lines of a conventional rectilinear k-space trajectory that are filled with the multiple echo-train acquisition after a single radiofrequency signal acquisition. Each 369
BAYRAMOGLU ET AL
Academic Radiology, Vol 17, No 3, March 2010
TABLE 2. Results of Image Data Analysis for Each Fat-suppressed T2W Sequence
Sequence Reader 1 T2W-PACEy-TSE with RP T2W-PACE-TSE BH-T2W-TSE with RP T2W-BLADEy with PACE and RP P Reader 2 T2W-PACE-TSE with RP T2W-PACE-TSE BH-T2W-TSE with RP T2W-BLADE with PACE and RP P
Motion Artifacts
Edge Sharpness of Liver, Pancreas, and Intrahepatic Vessels
Depiction of Intrahepatic Vessels
Overall Image Quality
3.88 0.52 3.74 0.48 3.24 0.43 4.84 0.47 *P2–P6 < .01 P1 > .05
4.12 0.43 4.00 0.50 3.40 0.50 4.92 0.40 *P2–P6 < .01 P1 > .05
4.20 0.50 4.00 0.64 3.52 0.50 4.88 0.33 *P1–P6 < .01
4.20 0.40 4.04 0.35 3.56 0.50 4.96 0.20 *P2–P6 < .01 P1 > .05
3.80 0.40 3.68 0.55 3.32 0.47 4.92 0.27 *P2–P6 < .01 P1 > .05
4.04 0.53 3.92 0.57 3.34 0.47 4.84 0.47 *P2–P6 < .01 P1 > .05
4.28 0.45 4.04 0.61 3.60 0.50 4.84 0.37 *P1–P6 < .01
4.24 0.43 4.04 0.45 3.52 0.50 4.88 0.43 *P1–P6 < .01
BH, breath-hold; P1, P value between first and second sequences; P2, P value between first and fourth sequences; P3, P value between first and third sequences; P4, P value between second and fourth sequences; P5, P value between second and third sequences; P6, P value between third and fourth sequences; PACE, Prospective Acquisition Correction; RP, restore pulse; TSE, turbo spin-echo; T2W, T2-weighted. Data are expressed as mean standard deviation. *P < .01. y Siemens Medical Systems (Erlangen, Germany).
blade was evenly rotated 11 times to cover the k-space completely. The width of each blade, which defines the number of phase-encoding lines per blade, was 35, and k-space coverage was 100%. The mean image acquisition times were 205 52 seconds for T2W-PACE-TSE with RP and T2W-PACE-TSE, 245 65 seconds for T2W-BLADE with PACE and RP, and 40 seconds for BH-T2W-TSE with RP. Image Analysis
In each subject, T2W-PACE-TSE, T2W-PACE-TSE with RP, BH-T2W-TSE with RP, and T2W-BLADE with RP images were acquired and evaluated qualitatively in terms of overall image quality, motion artifacts (ie, respiratory ghosting, vascular pulsation, peristalsis, and susceptibility), the depiction of the intrahepatic vessels, and the edge sharpness of the liver, pancreas, and intrahepatic vessels. In quantitative analysis, the mean signal intensities (SIs) of the liver, spleen, and gallbladder as well as background noise were measured, and the liver SNR (SIliver/SDnoise), liver-to-spleen contrast-to-noise ratio (CNR = [SIspleen SIliver]/SDnoise), and gallbladder SNR (SIgallbladder/SDnoise) were calculated, where SD is the standard deviation. All evaluated items were compared to one another statistically. Qualitative Evaluation
Assessments of images from all four sequences were performed by two independent readers with 4 and 5 years of experience in abdominal MRI. Both reviewers were unaware of the sequence parameters and patient information. The reviewers 370
graded the overall image quality of each T2W sequence using the following subjective five-point scale: 1 = unacceptable, 2 = poor, 3 = fair, 4 = good, and 5 = excellent. The presence of motion artifacts such as respiratory ghosting, vascular pulsation, peristalsis, and susceptibility was graded as 1 = severe, 2 = moderate, 3 = mild, 4 = minimal, or 5 = absent. The depiction of the intrahepatic vessels was assessed as 1 = unacceptable (invisible main portal vein), 2 = poor (only main portal vein visible), 3 = fair (only main branch of portal vein visible), 4 = good (some peripheral portal veins visible), or 5 = excellent (peripheral portal veins clearly visible). The edge sharpness of the liver, pancreas, and intrahepatic vessels was graded as 1 = unacceptable, 2 = poor, 3 = fair, 4 = good, or 5 = excellent. Quantitative Evaluation
Quantitative analysis was performed on the T2W magnetic resonance images with operator-defined region-of-interest (ROI) measurements of the mean SIs in the liver, spleen, and gallbladder as well as background noise. The SIs of the liver and spleen were measured in areas devoid of large vessels and prominent artifacts. The mean SIs of liver and spleen were determined with standard ROI measurements in three different sections from each set of images (the upper, middle, and lower parts of the spleen and liver). For all measurements, the size of the ROI was identical for all acquisition methods. The SD of background noise was measured in the largest possible ROI positioned in the phase-encoding direction outside the abdominal wall to account for any motion artifacts. The liver SNR, liver-to-spleen CNR, and gallbladder SNR were then calculated.
Academic Radiology, Vol 17, No 3, March 2010
T2W MRI OF THE UPPER ABDOMEN
Figure 1. Sample images of a patient with two hemangiomas in the liver: (a) T2-weighted (T2W) turbo spin-echo (TSE) Prospective Acquisition Correction (PACE); (b) T2W TSE PACE with restore pulse (RP); (c) T2W TSE breath-hold with RP; and (d) T2W BLADE PACE with RP.
Statistical Analysis
Statistical analysis was performed using NCSS 2007 and PASS 2008 statistical software (NCSS, Kaysville, UT). Friedman repeated-measures analysis of variance on ranks was used on the basis of negative results on tests assessing distribution normality. Wilcoxon’s signed-rank test was applied as a post hoc test for differences in multiple comparisons. In these analysis, P values < .05 were considered to indicate statistically significant differences. Reader agreement was assessed using Cohen’s k test. A value of 0 to 0.40 implied fair, 0.41 to 0.60 moderate, 0.61 to 0.80 substantial, and 0.81 to 1 almost perfect agreement. RESULTS Qualitative Analysis
Table 2 shows the results for motion artifacts; the edge sharpness of the liver, pancreas, and intrahepatic vessels; depictions of the intrahepatic vessels; and overall image quality as evaluated by the two observers for the four different sequences (Fig 1). The severity of motion artifacts induced by respiratory ghosting, vascular pulsation, peristalsis, and susceptibility was significantly lower in BLADE T2W TSE with RP images compared to the other sequences (P < .01). The edge sharpness of the liver, pancreas, and intrahepatic vessels and depictions of the intrahepatic vessels were also significantly better with this sequence (P < .05). There was no statistically significant difference between the T2W-PACE-TSE with RP and T2W-PACE-TSE sequences in terms of motion artifacts and the edge sharpness of the liver and pancreas. All examination criteria were significantly lower in the BH-T2W-TSE with RP sequence. Interobserver agreement for all evaluated items was judged almost perfect (Table 3).
TABLE 3. Results of Cohen’s k Statistical Analysis of Motion Artifacts, Edge Sharpness of the Liver and Pancreas, Depiction and Sharpness of Intrahepatic Vessels, and Overall Image Quality Between Readers Parameter Motion artifacts Edge sharpness of liver, pancreas, and intrahepatic vessels Depiction of intrahepatic vessels Overall image quality
k 88.2% 84.5% 86.6% 85.9%
Quantitative Analysis
The results of quantitative assessments of the mean background noise, liver SNR, liver-to-spleen CNR, and gallbladder SNR are summarized in Table 4. Differences among the sequences for these quantitative evaluation are shown in Table 5. The background noise with BH-T2W-TSE with RP was significantly lower than with the T2W-PACE-TSE with RP and BLADE sequences (P < .001). The mean liver SNR and liver-to-spleen CNR obtained with the BLADE sequence were significantly higher than those obtained using the other pulse sequences (P < .001). Similarly, the mean liver CNR was significantly higher with T2W-PACE-TSE with RP than with the T2W-PACE-TSE and BH-T2W-TSE with RP sequences. There was no statistically significant difference between the T2W-PACE-TSE and BH-T2W-TSE with RP sequences. The gallbladder SNR was highest with the BLADE sequence, but there was no statistically significant difference between the T2W-PACE-TSE with RP and BLADE. T2WPACE-TSE with RP was higher than the T2W-PACE-TSE and BH-T2W-TSE with RP sequences in terms of gallbladder 371
BAYRAMOGLU ET AL
Academic Radiology, Vol 17, No 3, March 2010
TABLE 4. Results of Quantitative Assessment of Background Noise, Liver SNR, Liver-to-Spleen CNR, and Gallbladder SNR Sequence T2W-PACE*-TSE with RP T2W-PACE-TSE T2W-BLADE* with PACE and RP BH-T2W-TSE with RP
Background Noise
Liver SNR
Liver-to-Spleen CNR
Gallbladder SNR
2.2 0.6 2.3 0.7 2.2 0.8 2.0 0.6
65.4 22.8 58.2 20.7 84.0 31.5 56.2 26.8
72.4 34.5 65.1 24.4 84.0 35.1 63.7 36.4
372.9 137.9 317.4 109.5 420.2 131.7 299.8 136.7
BH, breath-hold; CNR, contrast-to-noise ratio; PACE, Prospective Acquisition Correction; RP, restore pulse; SNR, signal-to-noise ratio; TSE, turbo spin-echo; T2W, T2-weighted. Data are expressed as mean standard deviation. *Siemens Medical Systems (Erlangen, Germany).
TABLE 5. Differences Between Sequences in Terms of Background Noise, Liver SNR, Liver-to-Spleen CNR, and Gallbladder SNR
Comparison T2W-BLADE* with PACE* and RP vs T2W-PACE-TSE with RP T2W-BLADE with PACE and RP vs T2W-PACE-TSE T2W-BLADE with PACE and RP vs BH-T2W-TSE with RP T2W-PACE-TSE with RP vs T2W-PACE-TSE T2W-PACE-TSE with RP vs BH-T2W-TSE with RP T2W-PACE-TSE vs BH-T2W-TSE with RP
Background Noise (P)
Liver SNR (P)
Liver-to-Spleen CNR (P)
Gallbladder SNR (P)
<.001 <.001 <.001 <.001 <.05 .740
<.001 <.001 <.001 <.001 <.05 .74
<.001 <.001 <.001 .07 <.05 .45
.08 <.01 <.001 <.001 <.001 .22
BH, breath-hold; CNR, contrast-to-noise ratio; PACE, Prospective Acquisition Correction; RP, restore pulse; SNR, signal-to-noise ratio; TSE, turbo spin-echo; T2W, T2-weighted. *Siemens Medical Systems (Erlangen, Germany).
SNR. There was no significant difference between the T2WPACE-TSE and BH-T2W-TSE with RP sequences in all quantitative evaluations. The mean image acquisition time for BH-T2W-TSE with RP was significantly shorter than for the other sequences (P < .001). There was no significant difference in the mean image acquisition time among other sequences. DISCUSSION Patient motion is a significant problem in MRI of the upper abdomen, leading to a reduction in image quality and a loss of diagnostic information. Motion caused by respiration and cardiac and vascular pulsation occurring during data acquisition causes image artifacts, a loss of resolution, and a reduction in SNR (26,27). It also may reduce anatomic details and lead to lesions’ being obscured in images. The BLADE technique offers advantages over other methods used for patient motion correction, because it uses inherent data to correct two main motions, in-plane rotation and translation, which occur in supine patients undergoing axial abdominal imaging (28). This is made possible by acquiring data in a series of concentric blades, each of which rotates through the center of the k-space. This offers two main benefits over conventional data collection. First, the central region of the k-space is sampled multiple times, and this method yields a better SNR by oversampling data at the center of the k-space. Second, data within this central region can be compared between each blade. If motion has occurred between the acquisition of each blade, data can be transposed to its 372
estimated stationary position, prior to the final image reconstruction. Therefore, artifact suppression is improved. Consequently, the BLADE sequence is less sensitive to motion than conventional sequences. Using the BLADE technique, image quality could be improved, especially in unsedated patients, without a loss of diagnostic information (21). In our study, we observed this sequence to provide the same advantages in upper abdominal images. There were significant differences among BLADE and the other sequences in terms of motion artifacts and image quality. We especially observed the depictions of intrahepatic vessels and the edge sharpness of solid organs to be superior with the BLADE sequence. The reason for this is a more uniform distribution of echo times throughout the k-space, resulting in improved image contrast. Several studies reporting improved visual image quality, reduced motion artifacts, and increased detectability of lesions in the brain and upper abdomen with PROPELLER (BLADE) sequences have been published (20–25,29). In our study also, the highest SNR, liver-to-spleen CNR, and gallbladder SNR were obtained with the BLADE sequence. This result shows that oversampling data at the center of the k-space yields a better SNR with the BLADE technique. On BLADE images, a radial artifact occasionally appeared around a structure with high SI (22). Our experience with the BLADE technique suggests that these artifacts are not a significant problem, and the radial dispersion of motion artifacts may serve to further reduce visible motion. A limitation of the BH T2W TSE sequence is decreased SI due to proton saturation with short repetition times (30).
Academic Radiology, Vol 17, No 3, March 2010
However, a modified TSE sequence using an additional RP at the end of multiple refocusing pulses was reported to provide excellent T2W images with higher lesion-to-liver contrast than the TSE sequence (7,18,19). On the other hand, several authors have suggested that respiratory-triggered and navigator-triggered TSE are better than BH TSE with RP (15). Although BH TSE with RP has been evaluated, this sequence has never been used with a navigator-triggered method. In our study, liver and gallbladder SNRs obtained with T2WPACE-TSE with RP were significantly higher than those obtained using T2W-PACE-TSE. The benefit of an additional RP in the TSE sequence is higher T2 contrast. In our study, this feature was accentuated especially in fluid-containing structures, such as the gallbladder. There are also several potential advantages of navigator-triggered TSE over the BH technique. T2W-PACE-TSE with RP was also significantly better than BH-T2W-TSE with RP in terms of SNR, liver-to-spleen CNR, and gallbladder SNR. The higher spatial resolution in the magnetic resonance images obtained with the navigatorecho technique using a larger matrix size and two signal acquisitions may be a possible reason for the superior image technique. Image sharpness was also significantly better with this technique than with the BH technique. PACE is used to improve the image quality and high-resolution examinations with thinner slice thicknesses. This is especially useful for patients who have difficulty undergoing BH examinations or those under sedation. Our study had one limitation. Because only nine solid lesions were found in our series, we did not evaluate detectability, although the detection of solid tumors is an important indicator for image contrast in abdominal MRI. Our goal was the evaluation of image quality. Hirokawa et al (31) reported that T2W MRI with PACE and PROPELLER techniques and superparamagnetic iron oxide enhancement is a promising method with which to improve the detection of hepatic lesions. Further studies should be performed to consider lesion detection in larger series. In conclusion, our findings suggest that BLADE MRI offers reductions in motion artifacts and improvements in image quality over conventional sequences. Using the PACE method improved image quality and allowed high-resolution examination with thinner slice thicknesses. An additional RP provides better image quality because of the acquisition of higher contrast.
REFERENCES 1. Tang Y, Yamashita Y, Namimoto T, et al. Liver T2-weighted MR imaging: comparison of fast and conventional half-Fourier single-shot turbo spinecho, breath-hold turbo spin-echo, and respiratory-triggered turbo spinecho sequences. Radiology 1997; 203:766–772. 2. Ito K, Mitchell DG, Outwater EK, et al. Hepatic lesions: discrimination of nonsolid, benign lesions from solid, malignant lesions with heavily T2weighted fast spin-echo MR imaging. Radiology 1997; 204:729–737. 3. Matsuo M, Kanematsu M, Murakami T, et al. T-weighted MR imaging for focal hepatic lesion detection: supplementary value of breath-hold imaging with half-Fourier single-shot fast spin-echo and multishot spin-echo echoplanar sequences. J Magn Reson Imaging 2000; 12: 444–452.
T2W MRI OF THE UPPER ABDOMEN
4. Pawluk RS, Borrello JA, Brown JJ, et al. A prospective assessment of breath-hold fast spin echo and inversion recovery fast spin echo techniques for detection and characterization of focal hepatic lesions. Magn Reson Imaging 2000; 18:543–551. 5. Hori M, Murakami T, Kim T, et al. Single breath-hold T2-weighted MR imaging of the liver: value of single-shot fast spin-echo and multishot spinecho echoplanar imaging. AJR Am J Roentgenol 2000; 174:1423–1431. 6. Augui J, Vignaux O, Argaud C, et al. Liver: T2-weighted MR imaging with breath-hold fast–recovery optimized fast spin-echo compared with breath-hold half-Fourier and non-breath-hold respiratory-triggered fast spin-echo pulse sequences. Radiology 2002; 223:853–859. 7. Yoshikawa T, Mitchell DG, Hirota S, et al. Focal liver lesions: breathhold gradient and spin-echo T2-weighted imaging for detection and characterization. J Magn Reson Imaging 2006; 23:520–528. 8. Herborn CU, Vogt F, Lauenstein TC, et al. MRI of the liver: can true FISP replace HASTE? J Magn Reson Imaging 2003; 17:190–196. 9. Reinig JW. Breath-hold fast spin echo MR imaging of the liver: a technique for high-quality T2-weighted images. Radiology 1995; 194:303–304. 10. Gaa J, Hatabu H, Jenkins RL, et al. Liver masses: replacement of conventional T2-weighted spin-echo MR imaging with breath-hold MR imaging. Radiology 1996; 200:459–464. 11. Yu JS, Kim KW, Kim YH, et al. Comparison of multishot turbo spin echo and HASTE sequences for T2-weighted MRI of liver lesions. J Magn Reson Imaging 1998; 8:1079–1084. 12. Pauleit D, Textor J, Bachmann R, et al. Improving the detectability of focal liver lesions on T2-weighted MR images: ultrafast breath-hold or respiratory-triggered thin section MRI? J Magn Reson Imaging 2001; 14:128–133. 13. Kanematsu M, Hoshi H, Itoh K, et al. Focal hepatic lesion detection: comparison of four fat-suppressed T2-weighted MR imaging pulse sequences. Radiology 1999; 211:363–371. 14. Zech CJ, Herrmann KA, Huber A, et al. High-resolution MR-imaging of the liver with T2-weighted sequences using integrated parallel imaging: comparison of prospective motion correction and respiratory triggering. J Magn Reson Imaging 2004; 20:443–450. 15. Lee SS, Byun JH, Hong HS, et al. Image quality and focal lesion detection on T2-weighted MR imaging of the liver: comparison of two high-resolution free-breathing imaging techniques with two breath-hold imaging techniques. J Magn Reson Imaging 2007; 26:323–330. 16. Klessen C, Asbach P, Kroencke TJ, et al. Magnetic resonance imaging of the upper abdomen using a free-breathing T2-weighted turbo spin echo sequence with navigator triggered prospective acquisition correction. J Magn Reson Imaging 2005; 21:576–582. 17. Katayama M, Masui T, Kobayashi S, et al. Fat-suppressed T2-weighted MRI of the liver: comparison of respiratory-triggered fast spin-echo, breath-hold single-shot fast spin-echo, and breath-hold fast-recovery fast spin-echo sequences. J Magn Reson Imaging 2001; 14:439–449. 18. Huang J, Raman SS, Vuong N, et al. Utility of breath-hold fast-recovery fast spin-echo T2 versus respiratory-triggered fast spin-echo T2 in clinical hepatic imaging. AJR Am J Roentgenol 2005; 184:842–846. 19. Akın O, Schwartz LH, Webler A, et al. Evaluation of focal liver lesions: fastrecovery fast spin-echo T2-weighted MR imaging. Clin Imaging 2006; 30: 322–325. 20. Forbes KP, Pipe JG, Bird CR, et al. PROPELLER MRI: clinical testing of a novel technique for quantification and compensation of head motion. J Magn Reson Imaging 2001; 14:215–222. 21. Forbes KP, Pipe JG, Karis JP, et al. 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 Am J Neuroradiol 2003; 24:794–798. 22. Wintersperger BJ, Runge VM, Biswas J, et al. Brain magnetic resonance imaging at 3 Tesla using BLADE compared with standard rectilinear data sampling. Invest Radiol 2006; 41:586–592. 23. 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; 15:986–995. 24. Kiryu S, Watanabe M, Kabasawa H, et al. Evaluation of super paramagnetic iron oxide-enhanced diffusion-weighted PROPELLER T2-fast spin echo magnetic resonance imaging. J Comput Assist Tomogr 2006; 30: 197–200. 25. Hirokawa Y, Isoda H, Maetani YS, et al. MRI artifact reduction and quality improvement in the upper abdomen with PROPELLER and Prospective Acquisition Correction (PACE) technique. AJR Am J Roentgenol 2008; 191:1154–1158.
373
BAYRAMOGLU ET AL
26. Bellon EM, Haacke EM, Coleman PE, et al. MR artifacts: a review. AJR Am J Roentgenol 1987; 148:1251–1258. 27. Low RN, Alzate GD, Schimakawa A. Motion suppression in MR imaging of the liver: comparison of respiratory-triggered and non-triggered fast spinecho sequences. AJR Am J Roentgenol 1997; 168:225–231. 28. Pipe JG. Motion correction with PROPELLER MRI: application to head motion and free-breathing cardiac imaging. Magn Reson Med 1999; 42: 963–969. 29. Hirokawa Y, Isoda H, Maetani YS, et al. Evaluation of motion correction effect and image quality with the periodically rotated overlapping parallel
374
Academic Radiology, Vol 17, No 3, March 2010
lines with enhanced reconstruction (PROPELLER) (BLADE) and parallel imaging acquisition technique in the upper abdomen. J Magn Reson Imaging 2008; 28:957–962. 30. Yamashita Y, Yamamato H, Namimoto T, et al. Phased array breath-hold versus non-breath-hold MR imaging of focal liver lesions: a prospective comparative study. J Magn Reson Imaging 1997; 7:292–297. 31. Hirokawa Y, Isoda H, Maetani YS, et al. Hepatic lesions: improved image quality and detection with the periodically rotated overlapping parallel lines with enhanced reconstruction technique-evaluation of SPIO-enhanced T2-weighted MR images. Radiology 2009; 251:388–397.