不同层厚螺旋CT造影对周围肺动脉血栓显示的比较

Evaluation of peripheral pulmonary embolism by different thickness of CTPA imagings

目的评价不同层厚CT肺动脉造影（CTPA）对周围肺动脉内肺栓塞栓子的显示情况，提高对周围性肺栓塞的诊断水平。方法分析2005年9月至2006年10月连续纳入的97例急性肺栓塞患者CTPA图像，以0．625mm层厚的原始横轴位图像为基准，在此基础上采用MIP技术重建1．25mm、2．5mm和5．0mm层厚的3组图像，比较4组不同层厚CTPA图像对周围肺动脉内肺栓塞栓子的显示有无差异；评价观察者之间对4组不同层厚图像上周围肺动脉内肺栓塞栓子显示的一致性。不同层厚组间肺动脉栓塞血管数之间比较采用卡方检验，组间观察一致性比较采用κ检验。结果0．625mm层厚原始轴位和1．25mm、2．5mm、5．0mm层厚MIP图像的平均图像数分别为403、201、101和53幅；在显示肺段动脉血栓方面，仅5．0mm层厚图像与另3组间有统计学差异（χ^2=60．099，P〈0．01），但各组图像的观察者间的一致性很好（κ值从0．751到0．992）；在显示亚段肺动脉血栓方面，0．625mm层厚与1．25mm层厚图像差异没有统计学意义（χ^2=1．051，P〉0．05），明显优于2．5mm和5．0mm层厚图像（χ^2值分别为151．892和204．553，均P〈0．01），但各组图像的观察者间的一致性较好（κ值为0．611～0．935）。结论重建1．25mm层厚的图像与0．625mm层厚的图像在显示周围肺动脉内的肺栓塞栓子方面具有同等的诊断效能，与此同时可以减少图像的重建数量和阅读时间。

Objective To evaluate whether there are difference among different thickness of CTPA imagings for detection of emboli in peripheral pulmonary arteries. Methods CTPA Iimages of 97 patients with acute pulmonary embolism were devided into four series for each patient： 0. 625 mm thick original transaxial soure images and 3 series of reconstructed images using the MIP technique with slab thicknesses of 1.25 mm,2. 5 mm,and 5. 0 mm. The difference was compared among 4 series of CTPA images for detection of emboli in peripheral pulmonary arteries, and the agreement of 2 observers for detection of emboli in peripheral pulmonary arteries in different thickness images was tested, Chi-Square test was taken for comparion of 4 series of CTPA images for detection of emboli in peripheral pulmonary arteries. Kappa test was taken for the agreement of 2 observers for detection of emboli in peripheral pulmonary arteries. Results The average images number of 0. 625 mm, 1.25 mm, 2. 5 mm, 5.0 mm thickness imaging was 403,201, 101, and 53, respectively. For detecting clots in segmental pulmonary arteries, there was statistically significant difference between 5.0 mm thick imaging and other 3 series imagings （ χ^2 = 60. 099, P 〈 0. 01 ）, but agreement of 2 observers in each serie imaging was excellent（ value of κ from 0. 751 to 0. 973 ）. For clots in subsegment pulmonary arteries, no statistical significant difference was observed between 0. 625 mm and 1.25 mm thick imagings（χ^2 = 1. 051 ,P 〉0. 05）, and superior to 2. 5 mm and 5.0 mm thick images（ χ^2 = 151. 892,204. 553, respectively,P 〈 0. 01 ） , but agreement of two observers in each serier imaging was good （ value of κ from 0. 611 to 0. 935 ）. Conclusion Compared with 0. 625 mm thick transaxial imaging, 1.25 mm thick MIP imaging is equally sensitve for detecting emboli in peripheral pulmonary arteries, but needs fewer images and less time for image reading.