Quantitative and Visual Benefits of Data-Driven Motion Correction on Oncologic PET/CT: A Prospective Cross-sectional...

Area of Science:

• Nuclear Medicine
• Medical Imaging
• Radiochemistry

Background:

• Conventional positron emission tomography (PET) respiratory gating uses a fraction of PET counts, potentially limiting image quality.
• Elastic motion correction with deblurring (EMCD) utilizes all PET counts for motion correction without increasing image noise.

Purpose of the Study:

• To evaluate the quantitative and visual effects of EMCD on FDG-PET and DOTATATE-PET images.
• To compare EMCD-corrected images against conventional optimal gate (OG) and ungated (UG) images.

Main Methods:

• Prospective study of 78 adults undergoing oncologic PET/CT with FDG or DOTATATE.
• Reconstruction of ungated (UG), belt-gating-derived optimal gate (BG-OG), belt-gating-EMCD (BG-EMCD), and data-driven gating-EMCD (DDG-EMCD) images.
• Quantitative analysis of lesion SUV-max and contrast-to-noise ratio (CNR), and visual assessment by two blinded readers.

Main Results:

• Motion-corrected images (BG-OG, BG-EMCD, DDG-EMCD) showed significantly higher lesion SUV-max than UG images (P<0.001).
• EMCD-based images (BG-EMCD, DDG-EMCD) demonstrated significantly higher CNRs compared to BG-OG and UG images (P<0.001).
• Readers preferred EMCD-based images over BG-OG images, with EMCD occasionally identifying more lesions.

Conclusions:

• EMCD-based motion correction significantly improves lesion quantitation, including higher SUVs and CNRs, and enhances overall image quality in oncologic PET.
• EMCD is a valuable tool for improving the quality of oncologic PET imaging on capable scanners.