非对比增强MR肺动脉血管成像诊断肺栓塞血流抑制反转时间的选择

Selection of blood flow suppression inversion time in unenhanced MR pulmonary angiography in the diagnosis of pulmonary embolism

目的:比较不同血流抑制反转时间(BSP-TI)条件下,非对比增强空间标记多反转脉冲序列(SLEEK)MR血管成像诊断肺栓塞(PE)的准确度并评价图像质量。方法:61例(男38例,女23例)CT肺动脉血管成像(CTPA)诊断为肺栓塞的患者,于CTPA检查后48h内行非对比增强SLEEK MRA检查。比较不同BSP-TI的SLEEK MRA图像质量,并以CTPA作为参考标准,计算并比较不同BSP-TI的SLEEK MRA诊断肺栓塞的准确度。结果:BSP-TI为900ms时SLEEK MRA图像质量最佳,不同BSP-TI(BSP-TI=500、700、900、1100、1300和1500ms)诊断肺栓塞的准确度分别为62.3%、89.3%、90.4%、84.6%、77.0%、69.3%;但当存在肺膨胀不全或肺部炎症时,BSP-TI为1300及1500ms时有助于显示该区域的肺动脉及栓子。结论:BSP-TI为900ms时大多数患者的SLEEK MRA图像质量最佳,诊断肺栓塞的准确度最高,但当存在肺膨胀不全或肺部炎症时,调高BSP-TI有助于该区域肺动脉及肺栓塞的显示。

Objective： To compare the accuracy of the diagnosis of pulmonary embolism （PE） with different blood flow suppression inversion time （BSP-TI） in unenhanced MR angiography by applying spatial labeling with multiple inversion pulses sequence （SLEEK） technique, and to evaluate the image quality as well. Methods： 61 patients （38 males and 23 females） with PE diagnosed by CT pulmonary angiography （CTPA） underwent unenhanced MR angiography with SLEEK technique within 48 hours after routine CTPA. Tlae image quality of SLEEK MRA witla different BAP-TI was compared. With CTPA as reference standard, the accuracy for PE detection with different BAP-TI in SLEEK MRA was calculated and compared. Results：The best image quality of SLEEK MRA was achieved in the condition of BSP-TI= 900ms. The diagnostic accuracy for PE with different BSP-TI （BSP-TI= 500ms, 700ms, 900ms, 1100ms, 1300ms, 1500ms） was 62.3 %, 89.3 %, 90.4%, 84.6 %,77.0 % and 69.3 %, respectively. However, when atelectasis or pulmonary infection was existed, PA and PE within these areas could be better displayed using BSP-TI= 1300ms and 1500ms. Conclusion： For most of the PE patients, the best image quality and the highest diagnostic accuracy of PE can be achieved using BSP-TI= 900ms, However,if atelectasis or pulmonary infection was existed,increasing BSP-TI could be helpful to show the pulmonary artery and embolus within those areas.