We previously developed a data-driven (DD) respiratory-motion (RM) correction method for conventional SPECT gamma-cameras (REGAT) and adapted it to the new CZT camera (Discovery NM 530c). We recently reported that RM correction with REGAT applied to CZT myocardial perfusion SPECT imaging (MPI) is clinically feasible and impacts substantially myocardial perfusion defects. In this evaluation, we study whether REGAT applied to MPI (Discovery NM 530c) is capable of generating a data-driven (DD) cardiac gating signal allowing the generation of valid global left ventricular (LV) function parameters (EDV: end diastolic volume; ESV: end systolic volume; EF: ejection fraction).

Were included 7 patients addressed for stress/rest MPI. All patients had prone stress MPI (2 MBq/Kg 99mTc-Tetrofosmin) and rest MPI 3-hours later (6 MBq/Kg). All acquisitions were made on Discovery NM 530c. Each acquisition was processed with REGAT to generate a dynamic SPECT acquisition study. The latter was processed to generate a DD cardiac gating signal and generate a mean DD cardiac GSPECT study (GSPECT-DD). In parallel, a mean ECG cardiac GSPECT study was generated using the ECG trigger signal provided by traditional ECG monitor (GSPECT-M). The 2 generated cardiac GSPECT studies were reconstructed on Xeleris workstation and processed with Emory Cardiac Toolbox (ECT). LV EDV, ESV and EF were compared between cardiac GSPECT-DD and GSPECT-M.

Stress LV EVD, ESV, and EF were 91±24mL, 29±13ml, and 68±10% vs 95±23ml, 29±12ml, and 70±11% with GSPECT-DD vs GSPECTM respectively (P:NS). Rest LV EDV, ESV, and EF were 97±21ml, 32±10ml, and 67±6% vs 101±21ml, 31±10ml, and 69±6% with GSPECT-DD vs GSPECT-M respectively (P:NS).

Data-driven cardiac gating of MPI with Discovery NM 530c processed with REGAT is clinically feasible. It provides LV global systolic function parameters similar to those provided by the traditional clinically used ECG monitor gating.