Indirect methods for improving parameter estimation of PET kinetic models

Abstract

Purpose: Parametric images obtained from kinetic modeling of dynamic positron emission tomography (PET) data provide a new way of visualizing quantitative parameters of the tracer kinetics. However, due to the high noise level in pixel-wise image-driven time-activity curves, parametric images often suffer from poor quality and accuracy. In this study, we propose an indirect parameter estimation framework which aims to improve the quality and quantitative accuracy of parametric images. Methods: Three different approaches related to noise reduction and advanced curve fitting algorithm are used in the proposed framework. First, dynamic PET images are denoised using a kernel-based denoising method and the highly constrained backprojection technique. Second, gradient-free curve fitting algorithms are exploited to improve the accuracy and precision of parameter estimates. Third, a kernel-based post-filtering method is applied to parametric images to further improve the quality of parametric images. Computer simulations were performed to evaluate the performance of the proposed framework. Results and conclusions: The simulation results showed that when compared to the Gaussian filtering, the proposed denoising method could provide better PET image quality, and consequentially improve the quality and quantitative accuracy of parametric images. In addition, gradient-free optimization algorithms (i.e., pattern search) can result in better parametric images than the gradientbased curve fitting algorithm (i.e., trust-region-reflective). Finally, our results showed that the proposed kernel-based post-filtering method could further improve the precision of parameter estimates while maintaining the accuracy of parameter estimates.

Reliable estimation of brain intravoxel incoherent motion parameters using denoised diffusion‐weighted MRI

“…Once the parametric coefficient image X is estimated, the filtered parametric image can be calculated by KX. For more details on the kernel‐based postfiltering method, refer to Huang, Liu and Lin …”

“…However, due to the fact that brain has a low D*/D ratio, the brain IVIM parametric images obtained from the proposed kernel‐based image denoising method may still exhibit high variance. To improve the precision of IVIM parameter estimates, we applied a kernel‐based postfiltering method to each IVIM parametric image. Basically, the kernel‐based postfiltering method used the Landweber algorithm to solve a linear equation, Y = KX, where K is the kernel matrix, X is the parametric coefficient image to be recovered and Y is the parametric image obtained from curve fitting.…”

“…The purpose of this study is to assess whether brain DW‐MRI data after denoising using the proposed kernel‐based method can provide a reliable estimation of IVIM perfusion parameters. In order to further improve the precision of IVIM parameter estimates, we use a kernel‐based postfiltering method that was developed to improve the precision of parametric imaging of dynamic positron emission tomography …”

“…In order to further improve the precision of IVIM parameter estimates, we use a kernel-based postfiltering method that was developed to improve the precision of parametric imaging of dynamic positron emission tomography. 17 2 | MATERIALS AND METHODS…”

Kernel-based curve-fitting method with spatial regularization for generation of parametric images in dynamic PET

“…Our recent study demonstrated that it is possible to improve the quality of PET parametric images by using three post-reconstruction techniques: image denoising, gradient-free curve-fitting of the temporal TACs and kernel-based post-filtering of the parametric images (Huang et al 2019). In particular, we found that the gradient-free pattern search (PatS) method (Audet and Dennis 2002) outperformed the gradient-based methods such as Levenberg-Marquardt (LM) (Levenberg 1944, Marquardt 1963) and trust-region-reflective (Byrd et al 1988, Branch et al 1999.…”