Novel Method to Estimate Kinetic Microparameters from Dynamic Whole-Body Imaging in Regular-Axial Field-of-View PET Scanners

Abstract

Background: For whole-body (WB) kinetic modeling based on a typical positron emission tomography (PET) scanner, a multipass multibed scanning protocol is necessary because of the limited axial field of view. Such a protocol introduces loss of early dynamics of the time-activity curve (TAC) and sparsity in TAC measurements, inducing uncertainty in parameter estimation when using prevalent least squares estimation (LSE) (i.e., common standard) especially for kinetic microparameters.

Purpose: We developed and investigated a method to estimate microparameters enabling parametric imaging, by focusing on general image qualities, overall visibility, and tumor detectability, beyond the common standard framework for fitting of data and parameter estimation.

Methods: Our parameter estimation method, denoted parameter combination-driven estimation (PCDE), has two distinctive characteristics: 1) improved probability of having one-on-one mapping between early and late dynamics in TACs (the former missing from typical protocols) at the cost of the precision of the estimated parameter, and 2) utilization of multiple aspects of TAC in selection of best fits. To compare the general image quality of the two methods, we plotted tradeoff curves for the normalized bias (NBias) and the normalized standard deviation (NSD). We also evaluated the impact of different iteration numbers of the ordered-subset expectation maximization (OSEM) reconstruction algorithm on the tradeoff curves. In addition, for overall visibility, a measure of the ability to identify suspicious lesions in WB (i.e., global inspection), the overall signal-to-noise ratio (SNR) and spatial noise (NSD_spatial) were calculated and compared. Furthermore, the contrast-to-noise ratio (CNR) and relative error of the tumor-to-background ratio (RE_TBR) were calculated to compare tumor detectability within a specific organ (i.e., local inspection). Furthermore, we implemented and tested the proposed method on patient datasets to further verify clinical applicability.

Results: With five OSEM iterations, improved general image quality was verified in microparametric images (i.e., reduction in overall NRMSE: 57.5, 71.1, and 56.1 [%] in the K₁, k₂, and k₃ images, respectively). The overall visibility and tumor detectability were also improved in the microparametric images. (i.e., increase in overall SNR: 0.2, 4.1, and 2.4; decrease in overall NSD_spatial: 0.2, 5.4, and 4.1; decrease in RE_TBR for a lung tumor: 17.5, 82.2, and 68.4 [%]; decrease in RE_TBR for a liver tumor: 255.8, 1733.5, and 80.3 [%], in K₁, k₂, and k₃ images, respectively; increase in CNR for a lung tumor: 1.3 and 1.0; increase in CNR for a liver tumor: 1.2 and 9.8, in K₁ and k₃ images, respectively). In addition, with five OSEM iterations, the differences in macroparametric images of the two methods were insignificant (i.e., overall NRMSE difference was within 10 [%]; differences in overall SNR, overall NSD_spatial, and CNRs for both tumors were within 1.0; and the difference in RE_TBR was within 10 [%] except for an exceptional case). For patient study, improved overall visibility and tumor detectability were demonstrated in micoparametric images.

Conclusions: The proposed method provides improved microkinetic parametric images compared to common standard in terms of general image quality, overall visibility, and tumor detectability.