Comparison of deep learning and particle smoother EM methods for estimation of Rb‐82 myocardial perfusion PET kinetic parameters

Abstract Background Positron emission tomography (PET) enables quantification of dynamic physiological processes through time‐resolved imaging. In myocardial perfusion PET, kinetic compartment modeling is used to estimate physiological parameters and derive myocardial blood flow. However, conventional nonlinear least squares (NLLS) estimation is sensitive to model misspecification when not all parameters can be reliably estimated and must instead be fixed or initialized using population averages, which can degrade accuracy. Purpose This work develops and evaluates two alternative kinetic analysis approaches for PET: a particle smoother‐based Expectation–Maximization method (PSEM) and a convolutional neural network (CNN). Methods Both methods were evaluated using simulated dynamic myocardial perfusion studies and compared against NLLS and a Kalman‐smoother‐based Expectation–Maximization (KEM) algorithm across multiple frame durations and noise levels. Results Across 2–10 s frames, the CNN achieved the lowest relative errors for all parameters (: 8.78%–4.98%, : 26.05%–25.50%, : 34.34%–22.76%), significantly outperforming NLLS, KEM, and PSEM (Holm‐adjusted at 1.0 noise, 2‐s frames), although performance degraded under out‐of‐distribution input‐function conditions. Conclusions Overall, the CNN provided the most accurate and robust in‐distribution kinetic parameter estimates across frame durations. In contrast, PSEM exhibited parameter‐dependent behavior, improving estimation while underperforming for , suggesting that further methodological refinement is needed.