Department of Radiology and Imaging, Hospital for Special Surgery, 535 E 70th St, New York, NY 10021.
Department of Biostatistics, Hospital for Special Surgery, New York, NY.
AJR Am J Roentgenol. 2019 Dec;213(6):W277-W286. doi: 10.2214/AJR.19.21606. Epub 2019 Aug 15.
The objective of our study was to compare the quality and diagnostic utility of the following three metal artifact reduction sequences in evaluating hip arthroplasties: conventional multiacquisition variable-resonance image combination selective (MAVRIC SL), isotropic MAVRIC SL, and reduced-TR isotropic MAVRIC SL. Ninety-three hip arthroplasties (85 total hip replacements and eight hip resurfacings [nine bilateral hips]) in 84 patients (38 men and 46 women; mean age ± SD, 69.1 ± 9.7 years old) were imaged and evaluated. A calibration scan determined the number of spectral bins needed for each implant, and isotropic and conventional MAVRIC SL images were acquired. Reduced-TR isotropic MAVRIC SL scans were acquired for 40 arthroplasties. Two board-certified radiologists blinded to MRI acquisition evaluated images for clinical and image quality features and compared images using a mixed-effects ordinal logistic regression model and odds ratios. Rater agreement was assessed with Gwet agreement coefficients. Scanning times were compared using mixed-effects linear regression. Significance was set at < 0.05. Calibration scans decreased the number of bins needed (median, 12 bins; interquartile range, 10-16 bins). Isotropic MAVRIC SL (mean scanning time, 7 minutes 16 seconds; 95% CI, 7 minutes 7 seconds-7 minutes 25 seconds) acquisitions had the longest scanning time, and conventional (mean, 5 minutes 46 seconds; 95% CI, 5 minutes 37 seconds-5 minutes 55 seconds) and reduced-TR isotropic (5 minutes 28 seconds; 95% CI, 5 minutes 15 seconds-5 minutes 41 seconds) MAVRIC SL acquisitions had scanning times that were similar. Both isotropic and reduced-TR isotropic MAVRIC SL images showed decreased blurring and improved visualization of the synovium and periprosthetic bone compared with conventional MAVRIC SL images ( < 0.001). Isotropic MAVRIC SL acquisitions more effectively improved signal-to-noise ratio (SNR), visualization of the synovium and periprosthetic bone, and lesion conspicuity and decreased blurring compared with reduced-TR isotropic MAVRIC SL acquisitions ( < 0.032). Isotropic MAVRIC SL acquisitions improve SNR, conspicuity of lesions, and visualization of synovium and periprosthetic bone and decrease blurring compared with conventional MAVRIC SL acquisitions.
传统多采集可变共振图像组合选择(MAVRIC SL)、各向同性 MAVRIC SL 和缩短 TR 各向同性 MAVRIC SL。84 名患者(38 名男性和 46 名女性;平均年龄±标准差,69.1±9.7 岁)的 93 个髋关节(85 个全髋关节置换和 8 个髋关节表面置换[9 个双侧髋关节])进行了成像和评估。校准扫描确定了每个植入物所需的光谱-bin 数量,并获取了各向同性和传统 MAVRIC SL 图像。40 个关节置换术获得了缩短 TR 的各向同性 MAVRIC SL 扫描。两位经过董事会认证的放射科医生对 MRI 采集进行了盲法评估,以评估图像的临床和图像质量特征,并使用混合效应有序逻辑回归模型和优势比比较图像。使用 Gwet 一致性系数评估评级者的一致性。使用混合效应线性回归比较扫描时间。显著性设置为<0.05。校准扫描减少了所需的-bin 数量(中位数,12 个-bin;四分位间距,10-16 个-bin)。各向同性 MAVRIC SL(平均扫描时间,7 分 16 秒;95%CI,7 分 7 秒-7 分 25 秒)采集的扫描时间最长,而传统(平均,5 分 46 秒;95%CI,5 分 37 秒-5 分 55 秒)和缩短 TR 的各向同性(5 分 28 秒;95%CI,5 分 15 秒-5 分 41 秒)MAVRIC SL 采集的扫描时间相似。与传统 MAVRIC SL 图像相比,各向同性和缩短 TR 各向同性 MAVRIC SL 图像均显示出模糊程度降低,并且滑膜和假体周围骨的可视化得到改善(<0.001)。与缩短 TR 的各向同性 MAVRIC SL 采集相比,各向同性 MAVRIC SL 采集更有效地提高了信噪比(SNR)、滑膜和假体周围骨的可视化以及病变的显著性和模糊程度降低(<0.032)。与传统 MAVRIC SL 采集相比,各向同性 MAVRIC SL 采集提高了 SNR、病变的显著性、滑膜和假体周围骨的可视化并减少了模糊。