Purpose: High-speed (HS) single-photon emission computed tomography (SPECT) with a recently developed solid-state camera shows comparable myocardial perfusion abnormalities to those seen in conventional SPECT. We aimed to compare HS and conventional SPECT images from multiple centres with coronary angiographic findings.
Methods: The study included 50 patients who had sequential conventional SPECT and HS SPECT myocardial perfusion studies and coronary angiography within 3 months. Stress and rest perfusion images were visually analysed and scored semiquantitatively using a 17-segment model by two experienced blinded readers. Global and coronary territorial summed stress scores (SSS) and summed rest scores (SRS) were calculated. Global SSS ≥3 or coronary territorial SSS ≥2 was considered abnormal. In addition the total perfusion deficit (TPD) was automatically derived. TPD >5% and coronary territorial TPD ≥3% were defined as abnormal. Coronary angiograms were analysed for site and severity of coronary stenosis; ≥50% was considered significant.
Results: Of the 50 patients, 13 (26%) had no stenosis, 22 (44%) had single-vessel disease, 6 (12%) had double-vessel disease and 9 (18%) had triple-vessel disease. There was a good linear correlation between the visual global SSS and SRS (Spearman's ρ 0.897 and 0.866, respectively; p < 0.001). In relation to coronary angiography, the sensitivities, specificities and accuracies of HS SPECT and conventional SPECT by visual assessment were 92% (35/38), 83% (10/12) and 90% (45/50) vs. 84% (32/38), 50% (6/12) and 76% (38/50), respectively (p < 0.001). The sensitivities, specificities and accuracies of HS SPECT and conventional SPECT in relation to automated TPD assessment were 89% (31/35), 57% (8/14) and 80% (39/49) vs. 86% (31/36), 77% (10/13) and 84% (41/49), respectively.
Conclusion: HS SPECT allows fast acquisition of myocardial perfusion images that correlate well with angiographic findings with overall accuracy by visual assessment better than conventional SPECT. Further assessment in a larger patient population may be needed to confirm this observation.