Geoffrey Soo MBBS, Abstract Co-Author: Nothing to Disclose

Kenneth K. Lau, Presenter: Nothing to Disclose

Multidetector CT pulmonary angiography (CTPA) is of increasing use as a diagnostic test for suspected pulmonary embolism (PE). Routine imaging protocols include multiplanar reconstructions with maximum intensity projection (MIP) technique to aid the detection of PE, at the cost of longer post-processing and reporting times. This retrospective study assesses the efficacy of multiplanar reconstruction CTPA images in PE detection. 

51 patients with previously reported pulmonary embolism on CTPA (using a 16-slice CT scanner with 3 mm reconstructed axial images, together with sagittal and coronal reconstructions using MIP technique) over a 3 month period were identified. A control group of 50 patients with previously reported negative CTPA studies for PE were added and the patients de-identified. Each reconstruction MIP series and axial images were reviewed in random order by 3 independent, blinded radiologists in separate sittings. All images were then reviewed by the 3 radiologists together to provide a consensus view as to the number of true emboli for comparison.

No PE's were detected in the control group. There was good inter-observer agreement for all imaging sequences (kappa scores 0.76 - 0.86).  237 emboli were detected by consensus. 2/237 emboli were missed on axial imaging, with no false positive results (98.7% sensitivity, 83.9% specificity). 157/237 emboli were detected using coronal reconstruction with 22 false postive and 11 false negative results (93.5% sensitivity, 81.5% specificity). Only 121/237 emboli were detected using sagittal reconstruction with 18 false positive and 15 false negative results. (89% sensitivity, 88% specificity). False positive results were often secondary to breathing/movement artefact or partial voluming artefact from adjacent anatomical structures (ie. lymph nodes or pulmonary veins). Dense intra-vascular contrast often masked adjacent PE on sagittal and coronal reconstruction due to MIP technique, with resulting false negatives.

Conventional axial imaging was superior to sagittal or coronal reconstruction in PE detection. Review of the sagittal and coronal reconstructions alone added little value to the PE detection rates.

Review of multiplanar MIP reconstruction alone is less accurate than axial imaging due to partial voluming artefact and masking effect by dense contrast within the vessel lumen from the MIP technique.

Soo, G, Lau, K, Multiplanar Reconstruction in CT Pulmonary Angiography: Can It Be Reviewed in Isolation? .  Radiological Society of North America 2008 Scientific Assembly and Annual Meeting, February 18 - February 20, 2008 ,Chicago IL. http://archive.rsna.org/2008/6006804.html