TY - JOUR T1 - <strong>Feasibility of standard and generalized Patlak Models for dynamic imaging of multiple organs using the uEXPLORER PET scanner</strong> JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 3185 LP - 3185 VL - 63 IS - supplement 2 AU - Fengyun Gu AU - Qi Wu AU - Jianmao Wu AU - Debin Hu AU - Tianyi Xu AU - Shuangliang Cao AU - Yun Zhou AU - Hongcheng Shi Y1 - 2022/06/01 UR - http://jnm.snmjournals.org/content/63/supplement_2/3185.abstract N2 - 3185 Introduction: Long axial field of view (LAFOV) PET scanners such as uEXPLORER can cover multiple organs simultaneously using a single-bed position. Patlak model has been applied for the total-body PET [18F]-FDG imaging with its simplicity in recent studies. But the performance of Patlak model is yet to be evaluated, especially for certain tissues like the liver where [18F]-FDG may exhibit mild positive uptake reversibility and bladder associated with the complex tracer excretion process. The objectives of this study are to investigate the feasibility of standard and generalized Patlak models for diverse tissues. Methods: The standard Patlak model (sPatlak) is engaged to estimate the slope - influx (Ki) using ordinary least square, whereby the transformed time activity curve is plotted against “normalized time” (Patlak et al. 1983). The sPatlak linear graphical analysis assumes an irreversible two tissue (2C) compartmental model. However, the assumptions may be broken by some tissues and the sPatlak plot is no longer linear. The non-linear generalized Patlak model (gPatlak) has the potential to address this issue, which introduces an additional exponential term characterized by the net efflux (Kloss) in addition to Ki (Patlak et al. 1985). This model is based on the reversible 2C model and solved by applying a basis function to linearize the estimation process (Karakatsanis et al. 2015). A 60-min dynamic scan was conducted on the uEXPLORER PET scanner after the administration of [18F] -FDG with the approval of the ethics committee of Zhongshan hospital and reconstructed as following sequences: 30×2s, 12×5s, 6×10s, 4×30s, 25×60s, 15×120s. A series of ROIs in diverse organs/tissues (liver, kidney, lung, bladder, bone, tumor, myocardium and brain) were drawn manually. The input function is obtained from a ROI over the descending aorta. sPatlak and gPatlak analyses are both implemented for each organ. The performance of these two models is evaluated by the goodness of fit. A common metric – R squared is employed and the value closer to 1 indicates a better linear fit. Ki (and Kloss) associated with sPatlak and gPatlak are also derived. Results: sPatlak and gPatlak plots with their best linear fits for eight organs are presented in Fig. 1. Corresponding R squared is also indicated in each plot. Results show that sPatlak and gPatlak both work very well for bone, tumor, myocardium and brain (R squared greater than 0.98). There is an apparent non-linear pattern on the sPatlak plot for bladder but the gPatlak plot presents a very strong linear relationship with R squared around 1. For other organs such as liver, kidney and lung, the R squared of sPatlak plot fit is between 0.85 and 0.95 and there is an acceptable linear trend. gPatlak analysis can improve the linear fits of these organs slightly with R squared from 0.91 to 0.97. Ki (and Kloss) values associated with sPatlak and gPatlak are reported in Table 1. For bone, tumor, myocardium and brain, Ki values are close from two approaches and Kloss can be neglected. Much higher Ki and nonnegligible Kloss with gPatlak analysis for other organs are presented. Conclusions: The performance of standard and generalized Patlak methods has been assessed for multiple organs from a model-fitting perspective. Standard Patlak is applicable to most of the organs except the bladder and generalized Patlak can bring some benefits for certain tissues with complex kinetic characteristics such as liver, kidney, lung, especially bladder. Further investigations in a patient cohort are ongoing. ER -