From the Divisions of Radiology.
Medical Physics in Radiology, German Cancer Research Center (DKFZ), Heidelberg.
Invest Radiol. 2019 Nov;54(11):704-711. doi: 10.1097/RLI.0000000000000596.
Diffusion-weighted imaging (DWI) is an important part of oncological magnetic resonance imaging (MRI) examinations, especially for tertiary cancer prevention in terms of early detection of recurrent disease. However, abdominal studies can be challenged by motion artifacts, poor signal-to-noise ratios, and visibility of retroperitoneal structures, which necessitates sequence optimization depending on the investigated region. This study aims at prospectively evaluating an adapted DWI sequence ex vivo and in vivo in oncologic patients undergoing abdominal MRI.
This institutional review board-approved, prospective study included phantom measurements, volunteer examinations, and oncologic patient examinations of the abdomen. Fifty-seven MRI examinations in 54 patients (mean age, 58 years; range, 21-90 years) were included into the analysis. The MRI examination were performed at a 1.5 T MRI scanner (MAGNETOM Aera; Siemens Healthcare, Erlangen, Germany) and included both a standard EPI-DWI (s-DWI; b = 50, 900 s/mm) and an adapted DWI (opt-DWI; EPI-DWI with b = 0, 50, 900, 1500 s/mm, acquisition with higher spatial resolution and optimized processing for the abdomen including motion correction, adaptive image combination, and background suppression). For b = 900 s/mm, the ratio of signal intensity in the normal tissue and the standard deviation of the noise in the surrounding air was quantitatively calculated; image quality and tissue differentiation parameters were rated by 2 independent, blinded readers using a 5-point Likert scale. Statistics included Wilcoxon signed-rank test and kappa statistic (P < 0.05/0.0125 after Bonferroni correction).
The DWI phantom demonstrated an optimized contour sharpness and inlay differentiation for opt-DWI. The apparent ratio of normal tissue signal/standard deviation of background noise at b = 900 s/mm of the right/left hemiabdomen was significantly increased in opt-DWI (mean, 71.9 ± 23.5/86.0 ± 43.3) versus s-DWI (mean, 51.4 ± 15.4/63.4 ± 36.5; P < 0.001). Image quality parameters (contour sharpness and tissue differentiation of upper abdominal and retroperitoneal structures) were significantly increased in opt-DWI versus s-DWI (P < 0.001). Interreader reliability test showed good agreement (kappa = 0.768; P < 0.001).
This study prospectively evaluated the potential of adapted DWI for screening in tertiary prevention of oncologic patients. An optimized DWI protocol with advanced processing achieved improved image quality in quantitative and qualitative analyses. Oncological optimization of DWI should be performed before its application in cancer patients to improve both screening and follow-up examinations, to better unleash the diagnostic potential of DWI.
扩散加权成像（DWI）是肿瘤磁共振成像（MRI）检查的重要组成部分，特别是在早期发现复发性疾病方面，对三级癌症预防具有重要意义。然而，腹部研究可能会受到运动伪影、信噪比差和腹膜后结构可见性的影响，因此需要根据研究区域优化序列。本研究旨在前瞻性评估一种改良的 DWI 序列在接受腹部 MRI 检查的肿瘤患者中的离体和体内表现。
本研究经机构审查委员会批准，为前瞻性研究，纳入了体模测量、志愿者检查和肿瘤患者的腹部检查。共纳入 54 例患者（平均年龄 58 岁；范围：21-90 岁）的 57 次 MRI 检查。MRI 检查在 1.5 T MRI 扫描仪（MAGNETOM Aera；西门子医疗，德国埃尔朗根）上进行，包括标准 EPI-DWI（s-DWI；b = 50、900 s/mm）和改良 DWI（opt-DWI；EPI-DWI，b = 0、50、900、1500 s/mm，采用更高的空间分辨率和针对腹部的优化处理，包括运动校正、自适应图像组合和背景抑制）。对于 b = 900 s/mm，定量计算正常组织的信号强度与周围空气噪声标准差的比值；2 位独立、盲法读者使用 5 分制 Likert 量表对图像质量和组织区分参数进行评分。统计学分析包括 Wilcoxon 符号秩检验和 Kappa 统计（Bonferroni 校正后 P < 0.05/0.0125）。
DWI 体模显示 opt-DWI 具有优化的轮廓锐度和镶嵌区分度。右/左半肝 b = 900 s/mm 时正常组织信号/背景噪声标准差的表观比值在 opt-DWI 中显著增加（平均值，71.9 ± 23.5/86.0 ± 43.3），而在 s-DWI 中（平均值，51.4 ± 15.4/63.4 ± 36.5；P < 0.001）。与 s-DWI 相比，opt-DWI 的图像质量参数（上腹部和腹膜后结构的轮廓锐度和组织区分度）显著提高（P < 0.001）。读者间可靠性测试显示具有良好的一致性（Kappa = 0.768；P < 0.001）。
本研究前瞻性评估了改良 DWI 用于肿瘤患者三级预防筛查的潜力。经过先进处理的优化 DWI 方案在定量和定性分析中均实现了更好的图像质量。在将 DWI 应用于癌症患者之前，应进行肿瘤学优化，以提高筛查和随访检查的质量，更好地发挥 DWI 的诊断潜力。
