1 All authors: Department of Radiology, Division of Thoracic Imaging, Massachusetts General Hospital, Harvard Medical School, 55 Fruit St, Founders Rm 202, Boston, MA 02114.
AJR Am J Roentgenol. 2014 Jul;203(1):118-22. doi: 10.2214/AJR.13.11842.
The objective of our study was to assess the effect of radiation dose reduction on the detection of pleural effusions, thickening, and calcifications.
Forty-five human cadavers (mean age at death, 60 ± 17 [SD] years; male-female ratio, 29:16; mean body mass index, 29 ± 5.7 [SD] kg/m(2)) were scanned at seven different dose levels (CT Dose Index volume [CTDIvol] = 20, 12, 10, 6, 4, 2, and 0.8 mGy) on a 128-MDCT unit (Definition FLASH). Images were reconstructed at a 3-mm slice thickness and 2-mm increment with filtered back projection (FBP) technique. Two chest radiologists independently reviewed all image series for the detection of pleural effusion, pleural calcification, and adjacent parenchymal opacification from atelectasis or consolidation. Objective image noise was measured at each dose level on the pleural effusion using ImageJ software. Data analysis was performed with the Student t test and kappa test.
Pleural effusions were seen in 39 of 45 cadavers on image series acquired at 2-20 mGy. Only 14 of 39 pleural effusions were identified at 0.8 mGy. Pleural effusions were not detected in 25 of 39 cadavers at 0.8 mGy because of photon starvation and increased image noise. Patient size was significantly larger in subjects with undetected pleural effusion than in those with detectable pleural effusion at 0.8 mGy (p < 0.01). Pleural calcifications and thickening (seen at 2-10 mGy images in three of three cadavers) were not identified on 0.8-mGy FBP images. On the other hand, adjacent parenchymal opacification could be assessed at all dose levels. The mean CT numbers of the pleural effusion were significantly lower on 0.8-mGy images than on images obtained at all other dose levels (-21 ± 55 [SD] vs 17.6 ± 19 HU, respectively) (p < 0.001).
Pleural effusions, thickening, and calcifications can be seen on FBP images reconstructed at a CTDIvol as low as 2 mGy (32-cm body phantom). CT at 0.8 mGy may provide suboptimal information on very small pleural effusions, pleural thickening, and calcifications.
本研究旨在评估降低辐射剂量对胸腔积液、增厚和钙化检测的影响。
在一台 128 层 MDCT 机(Definition FLASH)上,对 45 具人体尸体(死亡时的平均年龄为 60±17 岁[标准差];男女比例为 29∶16;平均体重指数为 29±5.7[标准差]kg/m2)进行了 7 个不同剂量水平(CT 剂量指数容积[CTDIvol]=20、12、10、6、4、2 和 0.8 mGy)的扫描。图像以 3-mm 层厚和 2-mm 层间距重建,采用滤波反投影(filtered back projection,FBP)技术。两位胸部放射科医生独立对所有图像系列进行了审查,以检测胸腔积液、胸腔钙化以及由肺不张或实变引起的相邻实质混浊。使用 ImageJ 软件在每个剂量水平上测量胸腔积液的客观图像噪声。采用 Student t 检验和 Kappa 检验进行数据分析。
在 2-20 mGy 采集的图像系列中,45 具尸体中有 39 具可见胸腔积液。仅在 0.8 mGy 时发现了 14 例胸腔积液。在 0.8 mGy 时,由于光子饥饿和图像噪声增加,39 例胸腔积液中有 25 例未被检测到。在 0.8 mGy 时未检测到胸腔积液的患者,其体型明显大于可检测到胸腔积液的患者(p<0.01)。在三具尸体的 2-10 mGy 图像中可见的胸腔钙化和增厚,在 0.8 mGy FBP 图像中未被识别。另一方面,相邻实质混浊可在所有剂量水平进行评估。胸腔积液的平均 CT 值在 0.8 mGy 图像上明显低于所有其他剂量水平(分别为-21±55[标准差]和 17.6±19 HU)(p<0.001)。
FBP 图像重建的 CTDIvol 低至 2 mGy(32-cm 体模)时,可显示胸腔积液、增厚和钙化。0.8 mGy 的 CT 可能无法提供关于非常小的胸腔积液、胸腔增厚和钙化的最佳信息。