Detection of pulmonary embolism using 16-slice multidetector-row computed tomography: evaluation of different image reconstruction parameters

To compare different image reconstruction parameters for detecting emboli of the pulmonary arteries according to anatomic levels using 16-slice multidetector-row computed tomography in patients suspected of having an acute pulmonary embolism (PE). Sixty-two patients (33 male and 29 female) with a clinically suspected acute PE were included in the present study. Multidetector-row computed tomography scans were performed using 16-mmx0.75-mm collimation. Based on the computed tomography data set, different image reconstruction parameters were used for each patient: axial slice thicknesses (STs) of 0.75, 2, 4, and 6 mm; axial maximum intensity projection (MIP) STs of 4 mm; and coronal STs of 2 and 4 mm. In joint reading fashion, 2 experienced radiologists reviewed examination findings regarding the presence and/or absence of a PE. The reference standard for visualization and detection of PEs was defined using the axial images with a 0.75-mm ST. In 23 of 62 patients, a PE was diagnosed. For main and lobar pulmonary arteries, the sensitivities and specificities were as follows: axial 2-mm images, 1.0/1.0; axial 4-mm images, 1.0/1.0; axial 6-mm images, 0.97/0.99; MIP 4-mm images, 0.95/0.99; coronal 2-mm images, 1.0/1.0; and coronal 4-mm images, 1.0/1.0. Regarding segmental and subsegmental pulmonary arteries, sensitivity and specificity varied: axial 2-mm images, 0.97/1.0; axial 4-mm images, 0.81/0.99; axial 6-mm images, 0.65/0.99; axial MIP 4-mm images, 0.63/0.99; coronal 2-mm images, 0.91/0.99; and coronal 4-mm images, 0.74/0.99. In detecting segmental and subsegmental PEs, only the axial images with an ST of 2 mm proved to have results comparable with the axial 0.75-mm images. Therefore, thin-slice collimation and ST are mandatory for visualization of segmental and subsegmental PEs in patients suspected of having an acute PE.