RESUMEN
BACKGROUND: We previously validated the use of a data-driven cardiac respiratory-motion (RM) correction method (REGAT) applicable to CZT SPECT myocardial perfusion imaging (MPI). In this study, we adapted the same process used with REGAT for RM to generate data-driven cardiac contraction triggers and corresponding cardiac contraction-gated SPECT studies (GSPECT-DD). We aimed to study its feasibility and compare its performances to GSPECT studies generated with ECG monitor-based triggers (GSPECT-ECG). METHODS: We included seven non-consecutive randomly chosen patients addressed for 1-day 99mTc-Tetrofosmin stress/rest MPI acquired with multi-pinhole CZT SPECT. We studied the degree of agreement between GSPECT-DD and GSPECT-ECG for the classification of acquired images into the 16 categories of mean cardiac cycle, and compared between the two methods the cine image quality and global LV systolic function of reconstructed studies. RESULTS: We found almost perfect agreement between cardiac contraction triggers generated with data-driven and ECG monitor-based methods. As compared to GSPECT-ECG, GSPECT-DD provided comparable and well-correlated LV global systolic function parameters and similar cine image quality at both stress and rest. CONCLUSIONS: Data-driven cardiac contraction gating using REGAT is feasible with low-dose and high-dose MPI CZT SPECT. It provides GSPECT-DD studies comparable to GSPECT-ECG.