Department of Radiology, Lucas Center for Imaging, Stanford University, 1201 Welch Rd, PS 055B, Stanford, CA 94305.
Santa Clara Valley Medical Center, San Jose, CA.
AJR Am J Roentgenol. 2021 Jun;216(6):1614-1625. doi: 10.2214/AJR.20.24172. Epub 2020 Aug 5.
Potential approaches for abbreviated knee MRI, including prospective acceleration with deep learning, have achieved limited clinical implementation. The objective of this study was to evaluate the interreader agreement between conventional knee MRI and a 5-minute 3D quantitative double-echo steady-state (qDESS) sequence with automatic T2 mapping and deep learning super-resolutionaugmentation and to compare the diagnostic performance of the two methods regarding findings from arthroscopic surgery. Fifty-one patients with knee pain underwent knee MRI that included an additional 3D qDESS sequence with automatic T2 mapping. Fourier interpolation was followed by prospective deep learning super resolution to enhance qDESS slice resolution twofold. A musculoskeletal radiologist and a radiology resident performed independent retrospective evaluations of articular cartilage, menisci, ligaments, bones, extensor mechanism, and synovium using conventional MRI. Following a 2-month washout period, readers reviewed qDESS images alone followed by qDESS with the automatic T2 maps. Interreader agreement between conventional MRI and qDESS was computed using percentage agreement and Cohen kappa. The sensitivity and specificity of conventional MRI, qDESS alone, and qDESS plus T2 mapping were compared with arthroscopic findings using exact McNemar tests. Conventional MRI and qDESS showed 92% agreement in evaluating all tissues. Kappa was 0.79 (95% CI, 0.76-0.81) across all imaging findings. In 43 patients who underwent arthroscopy, sensitivity and specificity were not significantly different ( = .23 to > .99) between conventional MRI (sensitivity, 58-93%; specificity, 27-87%) and qDESS alone (sensitivity, 54-90%; specificity, 23-91%) for cartilage, menisci, ligaments, and synovium. For grade 1 cartilage lesions, sensitivity and specificity were 33% and 56%, respectively, for conventional MRI; 23% and 53% for qDESS ( = .81); and 46% and 39% for qDESS with T2 mapping ( = .80). For grade 2A lesions, values were 27% and 53% for conventional MRI, 26% and 52% for qDESS ( = .02), and 58% and 40% for qDESS with T2 mapping ( < .001). The qDESS method prospectively augmented with deep learning showed strong interreader agreement with conventional knee MRI and near-equivalent diagnostic performance regarding arthroscopy. The ability of qDESS to automatically generate T2 maps increases sensitivity for cartilage abnormalities. Using prospective artificial intelligence to enhance qDESS image quality may facilitate an abbreviated knee MRI protocol while generating quantitative T2 maps.
潜在的缩短膝关节 MRI 的方法，包括使用深度学习进行前瞻性加速，其临床应用仍十分有限。本研究旨在评估常规膝关节 MRI 与 5 分钟 3D 定量双回波稳态（qDESS）序列（自动 T2 映射和深度学习超分辨率增强）的读者间一致性，并比较两种方法在关节镜手术结果方面的诊断性能。51 例膝关节疼痛患者行膝关节 MRI 检查，包括额外的 3D qDESS 序列，具有自动 T2 映射。傅立叶插值后，前瞻性深度学习超分辨率将 qDESS 切片分辨率提高两倍。一位肌肉骨骼放射科医生和一位放射科住院医生使用常规 MRI 对关节软骨、半月板、韧带、骨骼、伸肌机制和滑膜进行独立的回顾性评估。在为期 2 个月的洗脱期后，读者单独查看 qDESS 图像，然后查看带有自动 T2 图谱的 qDESS 图像。使用百分比一致性和 Cohen kappa 计算常规 MRI 和 qDESS 之间的读者间一致性。使用确切 McNemar 检验比较常规 MRI、单独 qDESS 和 qDESS 加 T2 图谱与关节镜检查结果之间的灵敏度和特异性。常规 MRI 和 qDESS 在评估所有组织时的一致性为 92%。所有影像学发现的 Kappa 值为 0.79（95%CI，0.76-0.81）。在 43 例行关节镜检查的患者中，常规 MRI（敏感性，58-93%；特异性，27-87%）和单独 qDESS（敏感性，54-90%；特异性，23-91%）之间的灵敏度和特异性差异无统计学意义（=0.23 至>0.99）））对软骨、半月板、韧带和滑膜。对于 1 级软骨病变，常规 MRI 的灵敏度和特异性分别为 33%和 56%；qDESS 为 23%和 53%（=0.81）；qDESS 加 T2 图谱为 46%和 39%（=0.80）。对于 2A 级病变，常规 MRI 的值分别为 27%和 53%，qDESS 为 26%和 52%（=0.02），qDESS 加 T2 图谱为 58%和 40%（<0.001）。前瞻性增强深度学习的 qDESS 方法与常规膝关节 MRI 具有很强的读者间一致性，在关节镜检查方面具有相当的诊断性能。qDESS 自动生成 T2 图谱的能力提高了对软骨异常的敏感性。使用前瞻性人工智能增强 qDESS 图像质量可能有助于简化膝关节 MRI 方案，同时生成定量 T2 图谱。
