Diagnosis of pulmonary embolism: planar images generated from V/Q SPECT are not a reliable substitute for traditional planar V/Q scan.

Abstract : OBJECTIVES: The use of summed planar images generated from single-photon emission computed tomography (SPECT) ventilation/perfusion (V/Q) scintigraphy has been proposed as a substitute for planar V/Q scans in order to use the revised PIOPED interpretation criteria when only SPECT acquisition is performed in patients with suspected pulmonary embolism. The aim was to evaluate the accuracy of angular summed planar scans in comparison with true planar images. METHODS: Patients included in the 'SPECT study' assessing the diagnostic performance of V/Q SPECT were analysed. Angular summed planar images were generated from SPECT acquisition data and compared with true planar scans. RESULTS: Angular summed planar images were successfully generated for 246 patients. Regarding interobserver variability, the interpretation result was different for 15 (6%) summed planar scans with an excellent degree of agreement (κ=0.92; 95% confidence interval 0.88-0.96). With regard to intermodality interpretation variability between conventional planar and angular summed images, the result was different for 63 (26%) of 246 patients with an intermodality degree of agreement of κ=0.66 (95% confidence interval 0.58-0.73). CONCLUSION: Planar images generated from SPECT V/Q scintigraphy are not a reliable substitute for true planar V/Q images.