Motion correction and its impact on quantification in dynamic total-body 18F-fluorodeoxyglucose PET

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Radiochemistry

Background:

• Total-body positron emission tomography (PET) offers unprecedented whole-body scanning capabilities.
• Dynamic PET scans are essential for kinetic information but are susceptible to patient motion.
• Motion degrades image quality and quantification accuracy in long dynamic PET acquisitions.

Purpose of the Study:

• To demonstrate a motion correction framework for dynamic total-body FDG PET imaging.
• To evaluate the importance of motion correction in improving image quality and quantification.
• To compare the proposed SyN-seq method with other correction techniques and uncorrected images.

Main Methods:

• A novel motion correction framework (SyN-seq) was developed using diffeomorphic registration of dynamic frames.
• Dynamic FDG PET scans from 12 subjects (7 healthy, 5 with tumors) were analyzed.
• The SyN-seq method sequentially aligned each frame to a reference position, parametrizing motion as diffeomorphisms.

Main Results:

• SyN-seq significantly improved visual quality in SUV images, especially in the brain and abdomen.
• Tumor SUVmean improved by an average of 5.35% (P=0.047) in subjects with tumors.
• Motion correction reduced inter-subject variability in K_i quantification by 11.8% and improved frame alignment.

Conclusions:

• Motion correction is vital for maintaining image quality and quantitative accuracy in dynamic total-body PET.
• The demonstrated SyN-seq framework effectively mitigates motion artifacts in dynamic total-body PET imaging.
• This correction method holds potential for applications like brain-gut axis and systemic disease imaging.