Patlak Parametric Image Estimation from Dynamic PET Using Diffusion Model Prior

TL;DR

This paper introduces a diffusion model-based framework for improving the quality of Patlak parametric images in dynamic PET by leveraging a learned prior and data consistency constraints, enhancing image estimation accuracy.

Contribution

It presents a novel diffusion model prior approach for kinetic modeling in dynamic PET, integrating static PET data and kinetic constraints for better parametric image quality.

Findings

Improved quality of Patlak parametric images demonstrated on total-body PET datasets.

The framework effectively incorporates static PET priors into dynamic image estimation.

Promising results suggest potential for clinical and research applications.

Abstract

Dynamic PET enables the quantitative estimation of physiology-related parameters and is widely utilized in research and increasingly adopted in clinical settings. Parametric imaging in dynamic PET requires kinetic modeling to estimate voxel-wise physiological parameters based on specific kinetic models. However, parametric images estimated through kinetic model fitting often suffer from low image quality due to the inherently ill-posed nature of the fitting process and the limited counts resulting from non-continuous data acquisition across multiple bed positions in whole-body PET. In this work, we proposed a diffusion model-based kinetic modeling framework for parametric image estimation, using the Patlak model as an example. The score function of the diffusion model was pre-trained on static total-body PET images and served as a prior for both Patlak slope and intercept images by leveraging their patch-wise similarity. During inference, the kinetic model was incorporated as a data-consistency constraint to guide the parametric image estimation. The proposed framework was evaluated on total-body dynamic PET datasets with different dose levels, demonstrating the feasibility and promising performance of the proposed framework in improving parametric image quality.