Automated movement correction for dynamic PET/CT images: evaluation with phantom and patient data.
Hu Ye1, Koon-Pong Wong1, Mirwais Wardak1
1Molecular and Medical Pharmacology, David Geffen School of Medicine at UCLA, Los Angeles, California, United States of America.
Plos One
|August 12, 2014
Summary
Head movement during PET/CT scans creates image artifacts. This study developed an automated movement correction (MC) method that significantly improved image quality and quantitative accuracy for dynamic FDDNP and FDG brain scans.
Area of Science:
- Nuclear Medicine
- Medical Imaging
- Neuroscience
Background:
- Head movement during dynamic brain PET/CT imaging causes misalignment between PET and CT data.
- This misalignment leads to artifacts in CT-based attenuation corrected PET images, impacting image quality and quantitative analysis.
- Parametric images derived from dynamic PET scans are particularly sensitive to these motion-induced artifacts.
Purpose of the Study:
- To develop and evaluate an automated retrospective image-based movement correction (MC) procedure for dynamic brain PET/CT imaging.
- To assess the impact of MC on image quality and quantitative accuracy of dynamic FDDNP and FDG PET/CT scans.
- To investigate the effectiveness of MC in reducing motion artifacts in patients with neurodegenerative diseases or poor compliance.
Main Methods:
- Developed an automated MC procedure involving CT-to-PET registration, CT-based attenuation correction, and PET frame re-alignment.
- Evaluated MC performance using a Hoffman phantom and dynamic FDDNP and FDG PET/CT scans from patients.
- Quantified changes in FDDNP distribution volume ratio (DVR) and FDG uptake constant (Ki) before and after MC using Logan analysis and image-derived input function.
Main Results:
- Phantom studies demonstrated high registration accuracy and improved PET image quality after MC.
- Patient studies revealed significant head movement (average displacement 6.92 mm) across all subjects, particularly in later PET frames.
- MC significantly reduced image artifacts and led to significant differences (P<0.05) in regional FDDNP DVR and FDG Ki values.
Conclusions:
- The automated MC procedure effectively corrects for head movement in dynamic brain FDDNP and FDG PET/CT scans.
- MC improves both the qualitative and quantitative aspects of dynamic brain PET/CT imaging.
- This method holds promise for enhancing diagnostic accuracy in neuroimaging studies utilizing dynamic PET/CT.
