13N-NH3 myocardial perfusion imaging with reduced scan duration: a feasibility study in the era of total-body PET/CT

Abstract

Purpose: To explore the feasibility of reducing scan duration of 13N-NH3 myocardial perfusion imaging (MPI) using a total-body PET/CT scanner. Methods: Forty-five patients with known or suspected coronary artery disease (CAD) performing rest 13N-NH3 MPI with total-body PET/CT were retrospectively included. PET data were acquired in list mode for 10 min, and reconstructed into sequence images of different scan duration: 10-min, 7-min, 5-min, 3-min, and 2-min (G10 to G2). Subjective visual evaluation including overall impression, image noise and lesion visibility was evaluated using 5-point Likert scale. Quantitative parameters including perfusion defect extent (Extent), total perfusion defect (TPD), summed rest score (SRS), end-diastolic volume (EDV), end-systolic volume (ESV), ejection fraction (EF), and myocardial blood flow (MBF) were analyzed. The full-time images (G10) were served as the reference. Results: There were no significant differences in subjective visual scores between G7-G5 and G10 groups (p > 0.05). A significant decrease in overall impression and image noise of G3-G2 was observed when compared to G10 (p < 0.05). However, no significant difference in lesion visibility was noted between G3 and G10 (p > 0.05). All G3 image quality was clinically acceptable (≥ 3 points). Except for EDV and ESV, other quantitative parameters showed no significant difference between G7-G3 and G10 (p > 0.05) and agreements were good (ICC = 0.974–0.998). For G2, only TPD exhibited no significant difference when compared to G10 (p > 0.05). Conclusion: Regarding imaging quality and parametric quantification accuracy of 13N-NH3 MPI, a 3-min scan is clinically acceptable, while a 5-min scan is sufficiently reliable.