The median scores of motion and visualization assessments for MR images without MultiVane were 2.0 and 2.0 for reviewer 1, and 1.5 and 1.5 for reviewer 2 (Table 3). The motion-induced artifacts degraded the image quality, overlapped bladder wall and tumor, and thus sometimes caused the misinterpretation of tumor location.
3.2 Improving image quality with MultiVane (PROPELLER)
The application of PROPELLER substantially reduced the artifacts induced by both respiration (Fig. 3) and peristalsis (Fig. 4). The improvement was seen in all cases with different levels of motion severity (Fig. 3). The peristalsis-induced artifacts were resolved on T2W images with MultiVane (Fig. 4). The median motion and visualization scores increased to 3 and 3.5 for reviewer 1, and 3.5 and 3.5 for reviewer 2. Bladder tumor margins were better delineated in all cases (Fig. 5). Nonparametric signed-rank test showed that both motion artifact and tumor visualization were significantly (both P < .001) improved. Both reviewers independently had this finding.
There was slight change in the signal contrast and morphologic shapes on T2W images with MultiVane. However, it had no impact on the diagnostic tasks performed in the study.
3.3 Reader preference bias and inter-observer variability
We found that the mean score difference for motion assessment without MultiVane (δ = 0.41) was the same as with MultiVane. That was also true for the visualization scoring where the δ values were 0.50 for both T2W data with and without MultiVane.
Out of 64 assessment scores from each reviewer, 24 (38%) were the same (absolute agreement), 26 (41%) revealed a difference of only 1 increment (0.5, strong agreement), 11 (17%) with a difference of 2 increments (1.0, intermediate agreement), and 3 (5%) with differences of 1.5 or 2 (little agreement). Scoring differences occurred predominately due to reader preference reflecting slightly more favorable scoring. This was found for example in a case when the first reviewer assigned a score of 3 (mild motion artifacts) while the second reviewer gave a score of 3.5 (little motion artifacts) for T2W images with MultiVane. The same case without MultiVane was scored 1 by the first reviewer and 1.5 by the second reader. Therefore, even though the preference bias resulted in κ scores of less than 0.2 for all assessments, the independent reviews still reached the agreement on the significant improvement of image quality (as detailed in the previous section).
A consensus review was performed to analyze the factors that contributed to the 3 score differences of 1.5 or 2.0. Differences occurred in cases in which there was only diffuse bladder wall thickening, instead of a bulky tumor. Some parts of the thickening were in the motion-affected areas, while other parts were clear of artifacts. One reviewer had focused on the whole bladder image quality while the other reader assigned scores based on the tumor image quality.
In pelvic imaging, many patients, due to their health condition, are unable to maintain a shallow regular breath during an MRI sequence. As a result, MR images of the bladder are often degraded by unavoidable breathing motion as the bladder is close to the superficial abdominal fat. Respiratory motion artifact was previously reported in pelvic imaging. There has not been any satisfactory noninvasive solution to this issue. Our data have shown that the application of PROPELLER can resolve respiratory motion artifacts of different severity.
Peristalsis is an involuntary motion of the GI track. Depending on patient anatomy, peristalsis may directly affect the bladder region on MRI. In this study, there were 2 scans in which peristalsis-induced artifacts were seen to disrupt the visualization of the bladder tumors. We have demonstrated that PROPPELLER acquisition could remove the artifact completely or substantially reduce it improved morphologic visualization.
Several studies have previously reported on the application of PROPELLER in pelvic MR imaging.[10,18] These studies, which mainly focused on the visualization of female organs and gynecological lesions, all concluded that PROPELLER could reduce artifacts and improve image quality. Nonetheless, the study by Fujimoto et al found no statistical significance between T2W with and without PROPELLER for the visualization of bladder organ. It is noteworthy that in this study bladder was imaged at 1.5T and not at the center of FOV. Meanwhile, our data with the focus on bladder imaging showed a significant improvement of bladder wall visualization using PROPELLER at 3T.
In theory, PROPELLER can correct any type of motion in MRI data by eliminating inconsistent sampled data due to patient motion. Two studies by Dietrich et al and Nagatomo et al concluded that PROPELLER could reduce the motion artifacts in anatomical shoulder MRI.[5,13] Bayramoglu et al and Hirokawa et al reported that PROPELLER was an effective approach to improve breathing motion artifacts for better image quality in T2W-MRI of the upper abdomen.[3,8] Other studies showed that the method also improved the anatomical imaging (T1W or/and T2W) of cardiac, lungs, and head and neck by motion reduction.[6,9,12,14] Our study is the first to report that the severity of respiration and peristalsis induced artifacts on bladder imaging can be compensated by using PROPELLER, which appears to be a valuable tool to ensure consistent diagnostic image quality.
Improvement of lesion or disease visualization is one of the most important clinical values of PROPELLER that has been reported. Meier-Schroers et al indicated that MultiVane could provide high image quality to improve the detection of lung lesions. In MRI of the female pelvis, Lane et al presented that imaging with PROPELLER was superior for delineation of ovarian borders and follicles as well as detection of fibroids. Eriksson et al showed that the application of PROPELLER had excellent tissue contrast in brain imaging and could visualize the internal hippocampal structures and tissue changes associated with hippocampal sclerosis. In our study, T2W imaging with PROPELLER showed substantial improvement in delineating bladder tumors and their margins.
Dietrich et al indicated that PROPELLER increased the scan time. Lane et al reported that PROPELLER only has a slightly longer scan time in female pelvic imaging. Hirokawa et al demonstrated that it is possible to reduce image artifacts and obtain better image quality with the same scan time and coverage with PROPELLER. Our results also indicated that the application of this k-space sampling technique can significantly reduce motion-induced artifacts and provides high-quality imaging of the morphology of the bladder as well as the pelvis while similar coverage, scan time and image contrast can still be achieved.
The utilization of a Likert scale in this study showed that reviewers had their own preferences for categorizing image quality that led to different baseline preference while scoring. However, the δ values and the case-by-case assessment showed that the reviewers applied their preferences consistently throughout the assessment of the data with or without MultiVane. Therefore, regardless of the inter-observer variability shown by kappa scores, both reviewers found a significant difference in the image quality improvement of P < .001 for both motion artifacts and tumor visualization. They both recommended the use of the PROPELLER technique for standard of care bladder MRI. Hence, it should be noted that while kappa values are often used to evaluate the inter-observer agreement, they may overestimate the variability with a Likert scale. As long as reviewers are consistent in their assessment, the preference bias will not prevent them from reaching the same conclusion as was seen with the significant improvement achieved with MultiVane in this study. The 3 large score differences (1.5 or 2) suggested that comprehensive instructions and training datasets should be used to minimize reader bias in relation to variation of anatomy and disease characteristics.
The study had several limitations. We noted a slight change in the signal contrast and morphologic shapes, and little random noise with the PROPELLER acquisition. However, the reviewers agreed that these changes did not affect the diagnosis and would be outweighed by the benefits of motion artifact reduction. Due to this slight, but distinctly visible change and the obvious reduction in motion artifacts, our reviewers pointed out that they always could identify which one was acquired with or without MultiVane (PROPELLER). We, therefore, concluded that randomization or not would not make a difference in the final results because the readers would have never been truly blinded. Thus, we decided to not anonymize the data and presented the data as we did. Our assessment included only 16 intra-individual bladder MRI comparisons in patients with identified motion on the conventional imaging. However, we consistently observed a substantial improvement in motion artifacts and tumor visualization in all cases with PROPELLER acquisition. With this observation, further studies will be conducted to assess the diagnostic improvement of PROPELLER in tumor staging and monitoring therapeutic response of bladder cancer using pathological findings as a reference standard.
The PROPELLER data acquisition significantly reduced motion artifacts and thereby improved the visualization of bladder tumors. We propose that clinical T2W MRI of the bladder should utilize this motion compensation approach to ensure consistent image quality for data assessment.
Conceptualization: Huyen Thanh Nguyen, Michael Vinzenz Knopp.
Formal analysis: Huyen Thanh Nguyen, Zarine Ketul Shah, Michael Vinzenz Knopp.
Funding acquisition: Michael Vinzenz Knopp.
Investigation: Huyen Thanh Nguyen, Zarine Ketul Shah, Amir Mortazavi, Kamal S. Pohar, Michael Vinzenz Knopp.
Methodology: Huyen Thanh Nguyen, Zarine Ketul Shah, Amir Mortazavi, Kamal S. Pohar, Lai Wei, Debra Lyn Zynger, Michael Vinzenz Knopp.
Project administration: Huyen Thanh Nguyen.
Resources: Huyen Thanh Nguyen, Zarine Ketul Shah, Amir Mortazavi, Kamal S. Pohar, Lai Wei, Debra Lyn Zynger, Michael Vinzenz Knopp.
Software: Huyen Thanh Nguyen.
Supervision: Huyen Thanh Nguyen, Zarine Ketul Shah, Amir Mortazavi, Kamal S. Pohar, Michael Vinzenz Knopp.
Validation: Huyen Thanh Nguyen, Zarine Ketul Shah, Michael Vinzenz Knopp
Visualization: Huyen Thanh Nguyen, Zarine Ketul Shah.
Writing – original draft: Huyen Thanh Nguyen.
Writing – review and editing: Huyen Thanh Nguyen, Zarine Ketul Shah, Amir Mortazavi, Kamal S. Pohar, Lai Wei, Debra Lyn Zynger, Michael Vinzenz Knopp.
Huyen Thanh Nguyen orcid: 0000-0002-4409-0824.
Huyen Thanh Nguyen orcid: 0000-0002-4409-0824.
. Griffiths TR. Action on Bladder Cancer
. Current perspectives in bladder cancer
management. Int J Clin Pract 2013;67:43548.
. Zand KR, Reinhold C, Haider MA, et al. Artifacts and pitfalls in MR imaging of the pelvis. J Magn Reson Imaging 2007;26:48097.
. Bayramoglu S, Kilickesmez O, Cimilli T, et al. T2-weighted MRI of the upper abdomen: comparison of four fat-suppressed T2-weighted sequences including PROPELLER (BLADE) technique. Acad Radiol 2010;17:36874.
. Chen X, Xian J, Wang X, et al. Role of periodically rotated overlapping parallel lines with enhanced reconstruction diffusion-weighted imaging in correcting distortion and evaluating head and neck masses using 3 T MRI. Clin Radiol 2014;69:4039.
. Dietrich TJ, Ulbrich EJ, Zanetti M, et al. PROPELLER technique to improve image quality of MRI of the shoulder. AJR Am J Roentgenol 2011;197:W1093100.
. Eriksson SH, Thom M, Bartlett PA, et al. PROPELLER MRI visualizes detailed pathology of hippocampal sclerosis. Epilepsia 2008;49:339.
. 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;14:21522.
. Hirokawa Y, Isoda H, Maetani YS, et al. 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;28:95762.
. Huang TY, Tseng YS, Tang YW, et al. Optimization of PROPELLER reconstruction for free-breathing T1-weighted cardiac imaging. Med Phys 2012;39:4896902.
. Lane BF, Vandermeer FQ, Oz RC, et al. Comparison of sagittal T2-weighted BLADE and fast spin-echo MRI of the female pelvis for motion artifact and lesion detection. AJR Am J Roentgenol 2011;197:W30713.
. Mahmoud OM, Tominaga A, Amatya VJ, et al. Role of PROPELLER diffusion-weighted imaging and apparent diffusion coefficient in the evaluation of pituitary adenomas. Eur J Radiol 2011;80:4127.
. Meier-Schroers M, Kukuk G, Homsi R, et al. MRI of the lung using the PROPELLER technique: artifact reduction, better image quality and improved nodule detection. Eur J Radiol 2016;85:70713.
. Nagatomo K, Yabuuchi H, Yamasaki Y, et al. Efficacy of periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) for shoulder magnetic resonance (MR) imaging. Eur J Radiol 2016;85:173543.
. Ohgiya Y, Suyama J, Seino N, et al. MRI of the neck at 3 Tesla using the periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) (BLADE) sequence compared with T2-weighted fast spin-echo sequence. J Magn Reson Imaging 2010;32:10617.
. Pipe JG. Motion correction with PROPELLER MRI: application to head motion and free-breathing cardiac imaging. Magn Reson Med 1999;42:9639.
. Pipe JG, Farthing VG, Forbes KP. Multishot diffusion-weighted FSE using PROPELLER MRI. Magn Reson Med 2002;47:4252.
. Tamhane AA, Arfanakis K. Motion correction in periodically-rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) and turboprop MRI. Magn Reson Med 2009;62:17482.
. Fujimoto K, Koyama T, Tamai K, et al. BLADE acquisition method improves T2-weighted MR images of the female pelvis compared with a standard fast spin-echo sequence. Eur J Radiol 2011;80:796801.
Keywords:Copyright © 2019 the Author(s). Published by Wolters Kluwer Health, Inc.
bladder cancer; motion-induced artifacts; PROPELLER acquisition; T2W-MRI