Absolute Myocardial Blood Flow via Single-Time-Point Imaging in PET and SPECT: Cross-Validation Against Dynamic 13N-Ammonia PET.
Abstract
Background
Quantitative assessment of myocardial blood flow (MBF) and flow reserve (MFR) is clinically superior to relative imaging but limited by dynamic scanning requirements. We evaluated a novel single-time-point (STP) method to estimate MBF from a single static image using a population-based power law model of the arterial input function.
Methods
The STP model was optimized using retrospective dynamic 13N-Ammonia PET/CT data (n = 90) and prospectively validated in 31 symptomatic patients undergoing both dynamic 13N-Ammonia PET (reference) and conventional 99mTc-MIBI SPECT/CT (leave-one-patient-out cross-validation). STP estimates were compared against dynamic PET via correlation and Bland-Altman analyses.
Results
The STP model demonstrated stable performance across a wide post-injection window in PET. For 99mTc-MIBI SPECT, reproducibility coefficients for stress MBF (44%), rest MBF (58%) and MFR (67%) were comparable to PET test-retest literature. SPECT MBF 95% limits of agreement (range: 1.5 mL/g/min) approached or exceeded reported dynamic CZT-SPECT benchmarks. Technical acquisition delays were the primary source of model degradation.
Conclusion
The simplified STP approach provides reliable MBF and MFR estimates from standard static scans, showing strong agreement with dynamic 13N-Ammonia PET. By eliminating the requirement for full dynamic protocols, this method facilitates absolute flow quantification using existing conventional SPECT infrastructure.