Comparative assessment of linear least-squares, nonlinear least-squares, and Patlak graphical method for regional and local quantitative tracer kinetic modelling in cerebral dynamic ¹⁸ F-FDG PET

摘要： Purpose: Dynamic 18F-FDG PET allows quantitative estimation of cerebral glucose metabolism both at the regional and local (voxel) level. Although sensitive to noise and highly computationally expensive, nonlinear least-squares (NLS) optimization stands as the reference approach for the estimation of the kinetic model parameters. Nevertheless, faster techniques, including linear least-squares (LLS) and Patlak graphical method, have been proposed to deal with high resolution noisy data, representing a more adaptable solution for routine clinical implementation. Former research investigating the relative performance of the available algorithms lack precise evaluation of kinetic parameter estimates under realistic acquisition conditions. Methods: The present study aims at the systematic comparison of the feasibility and pertinence of kinetic modelling of dynamic cerebral 18F-FDG PET using NLS, LLS, and Patlak method, based on numerical simulations and patient data. Numerical simulations were used to study and parameters estimation bias and variance under representative noise levels. Patient data allowed to assess the concordance between the three methods at the regional and voxel scale, and to evaluate the robustness of the estimations with respect to patient head motion. Results and Conclusions: Our findings indicate that at the regional level NLS and LLS provide kinetic parameter estimates ( and ) with similar bias and variance characteristics ( bias ± rel. std dev. 0.0±5.1% and 0.1%±4.9% for NLS and LLS respectively, bias ± rel. std dev. 0.1%±4.5% and -0.7%±4.4% for NLS and LLS respectively), NLS estimates being however slightly less sensitive to patient motion. At the voxel level, provided that patient motion is negligible or corrected, LLS offers an appealing alternative solution for local mapping, with high correlation with NLS values (Pearson’s r = 0.95 on actual data) in computations times less than two orders of magnitude lower. Last, Patlak method appears as the most robust and accurate technique for the estimation of values at the regional and voxel scale, with or without head motion. It provides low bias / low variance quantification (bias ± rel. std dev. -1.5±9.5% and -4.1±19.7% for Patlak and NLS respectively) as well as smooth parametric images suitable for visual assessment.