Dynamic PET image reconstruction integrating temporal regularization associated with respiratory motion correction...

Area of Science:

• Nuclear Medicine
• Medical Imaging Physics
• Oncology Imaging

Background:

• Respiratory motion degrades PET image quality, impacting both qualitative and quantitative accuracy.
• Dynamic PET imaging for kinetic modeling is particularly sensitive to motion-induced statistical limitations and reduced temporal resolution.

Purpose of the Study:

• To develop and evaluate a novel reconstruction algorithm for dynamic PET imaging that integrates respiratory motion correction and temporal regularization.
• To address the limitations of standard reconstruction methods in dynamic PET for oncology applications.

Main Methods:

• A unique reconstruction algorithm combining elastic motion correction (estimated from non-attenuation-corrected PET) with temporal regularization using basis functions.
• List-mode OSEM reconstruction incorporating simultaneous estimation of temporal basis functions and coefficients.
• Quantitative evaluation using dynamic FDG PET/CT in lung cancer patients and simulated data.

Main Results:

• The proposed 4D algorithm significantly reduces noise and compensates for motion-induced contrast loss.
• Simulated data showed up to 40% bias reduction in tumor and blood regions with comparable standard deviations versus 3D reconstruction.
• Improved Patlak parameter estimation, with up to 40% bias reduction for slope and 30% for intercept in tumor regions.

Conclusions:

• Integrating respiratory motion correction and temporal regularization in dynamic PET reconstruction substantially enhances quantitative accuracy and reduces motion artifacts.
• The developed method shows significant potential for improving diagnostic and therapeutic assessments in oncology using dynamic PET imaging.