Impact of Equilibration Time (t*) on Patlak Quantitation in Dynamic Total-Body Imaging using the uEXPLORER PET Scanner

Introduction: As the most popular graphical method, Patlak model has been widely applied to dynamic PET imaging due to its simplicity and robustness. It is assumed that Patlak plot results in a straight line after equilibration time (t*) for systems with irreversible compartments. There is no doubt that the violation of assumption could introduce additional errors in estimated influx (Ki), such as the improper t*. With the advent of total body PET scanners like uEXPLORER, multiple organs can be covered simultaneously using a single-bed position. When adopting Patlak method for dynamic total-body PET imaging on such scanners, some new opportunities and challenges also emerge, for example, the single t* may not be appropriate for diverse tissues. The objectives of this study are to assess the performance of several t* for 18F-FDG studies in the published literature and propose an adaptive t* scheme for diverse tissues to enhance the quantitation.

Methods: Three fixed t*: 10 min, 20 min and 30 min reported in previous total-body studies are applied to implement Patlak analysis. An adaptive t* scheme considering the feasibility for multiple organs is also developed. The choice of optimal t* is based on two criterions - Max-Error and R squared, separately. Max-Error is defined as the worst case error between the predicted value and the true value for all observations on Patlak plot. The selected t* is the earliest one so that Max-Error is less than a threshold value. This criterion has been employed in PMOD and the default setting of threshold is 10%. R squared is a common metric to quantify the goodness of linear fit and a value closer to 1 indicates a better fit, so optimal t* is determined by the maximum R squared. The performance of t* is evaluated using a lung cancer patient study at regions of interest (ROIs) and voxel levels. A 60-min dynamic scan was performed on the uEXPLORER PET scanner immediately 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, spleen, kidney, lung, bladder, bone, tumour, myocardium and brain) were drawn manually. The input function is obtained from an ROI over descending aorta. Ki values for each ROI and parametric images are derived from Patlak analysis with three fixed and two adaptive t* determined by the above two criterions.

Results: Two metrics – Max-Error and R squared across all possible t* between 5 and 40 min are shown (Fig. 1). The optimal t* for each ROI based on these criterions is also indicated. Comparisons of the linear fitting among three fixed t*(10, 20 and 30 min) and two adaptive t* are presented (Fig. 2). Results show t* has a considerable impact on some specific organs like liver, kidney and bladder. One notable point is the apparent non-linear pattern on the Patlak plot for bladder. Ki values associated with these t* are calculated and reported in Table 1. Ki decreases with the increase of t* for most of organs, and the differences of Ki for spleen, lung, bone, tumor, myocardium and brain with two adaptive t* are not substantial – within 13%. Use either adaptive one as the reference, 10 min is the best among three fixed t*. Parametric images of Ki associated with different t* are also presented (Fig. 3).

Conclusions: The liver and kidney are sensitive to t* and Patlak model is not feasible for bladder. 10 min is an appropriate choice for most of the organs in this study. The proposed adaptive scheme has the potential to improve the accuracy of kinetic parameters such as Ki. Further investigations in a patient cohort and more sophisticated techniques are under development.

• Download figure
• Open in new tab
• Download powerpoint