TY - JOUR T1 - <strong>Using a Patlak plot-based optimization approach to estimate input function with incomplete blood samples for quantification of nonhuman primate dynamic 18F-FDG PET</strong> JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 582 LP - 582 VL - 60 IS - supplement 1 AU - Rongfu Wang AU - Xueqi Chen AU - Sulei Zhang AU - Jianhua Zhang AU - Lixin Chen AU - Yun Zhou Y1 - 2019/05/01 UR - http://jnm.snmjournals.org/content/60/supplement_1/582.abstract N2 - 582Objectives: FDG uptake rate constant Ki is a main physiology parameter to be measured by quantitative dynamic PET study. The Ki is a macro-parameter which can be estimated by fitting a compartmental model to the tissue kinetics measured by dynamic PET. Due to its simplicity, a model-independent to graphical analysis using Patlak plot is commonly used to estimate Ki, especially for total-body dynamic PET study. The objective of the study is to evaluate the Patlak plot-based optimization approach to estimate input function with incomplete blood samples for quantification of dynamic FDG PET. Methods: Eight 60-min monkey dynamic FDG-18F PET studies with total arterial blood samples were collected. The measured plasma input function (mPIF) was determined by as much as 34 arterial blood samples during each PET scan. Time activity curves (TACs) of 7 cerebral regions of interests (ROIs) were generated from each study. With given limited number of blood sampling, the optimal time points for those blood samples to estimate PIF (ePIF) was determined by maximizing the correlations between the Ki estimated ePIF and ones estimated mPIF. The estimated PIF (ePIF) from the incomplete sampling data was determined by either interpolation or extrapolation method using scale calibrated population mean of normalized PIF. A leave-two-out cross validation method was used to generate population mean of PIF. The Patlak plot was applied ROI TACs to estimate FDG-18F uptake rate constant Ki. Results: The linear correlations between the Ki estimates from ePIF with optimal sampling schemes and those from measured PIF were: Ki (ePIF 1 sample at 40 min) = 1.09 Ki (mPIF) - 0.00, R2 =0.95±0.08 ; Ki (ePIF 2 samples at 35 and 50 min) = 1.09 Ki (mPIF) - 0.00, R2 =0.95±0.07; Ki (ePIF 3 samples at 12, 40, and 50 min) = 1.04 Ki (mPIF) - 0.00, R2 =0.96±0.05; and Ki (ePIF 4 samples at 10, 20, 40, and 50 min) = 1.02 Ki (mPIF)-0.00, R2 =0.97±0.04. As sample size became greater or equal 4, the Ki estimates from ePIF with its corresponding optimal sampling protocol were almost identical to those from mPIF. Conclusions: The Patlak plot-based optimization approach is a robust method to estimate PIF from incomplete blood samples for quantification of non-human primate dynamic 18F-FDG PET, and this method could be potentially applied for quantification of human dynamic 18F-FDG PET studies. Keywords: Quantitative dynamic PET; Input function; Graphical analysis; Monkey; Patlak plot ER -