Abstract
2025
Objectives Single Photon Emission Computed Tomography (SPECT) has acquired popularity in ventilation / perfusion (VQ) scanning (1). Some centres acquire both SPECT and planar images due to unfamiliarity with reporting SPECT. This increases imaging time and decreases patient comfort. A solution is to generate 2D images from the 3D SPECT dataset, which resemble traditional planar scans (2). Initial reports are in disagreement with regard to the diagnostic accuracy of this method (3,4). We sought to compare the diagnostic interpretation of traditional VQ planar images with planar-like images reprojected from SPECT.
Methods Retrospective data from patients who had both a planar and a SPECT acquisition were used to generate anonymised reprojected planar images and were compared to traditional planar V/Q images. Two consultants blindly interpreted both sets of images for 81 patients following a proforma. We assessed the agreement in final diagnosis between the two imaging methods as well as between the two clinicians using contigency tabulation and Cohen's kappa coefficient. We also compared number, nature and localisation of defects, as well as image quality. Finally, we compared the diagnosis using planar methods to the original diagnosis that was made using SPECT.
Results There was excellent agreement in diagnosis both between the two planar methods (κ=0.93) (Table 1) and between the two consultants (κ=0.91). Number of defects, localisation and image quality were similar for the two imaging methods. Six additional PEs were diagnosed using 3D SPECT data.
Conclusions We have shown that reprojected planars from SPECT V/Q performed similarly to traditional planars. These results have potential important implications for patient workflow in busy Nuclear Medicine departments as well as for patient comfort.