Clinical Evaluation of a Data-Driven Respiratory Gating Algorithm for Whole-Body PET with Continuous Bed Motion

Area of Science:

• Nuclear Medicine
• Medical Imaging
• Radiochemistry

Background:

• Respiratory gating is standard in PET to mitigate motion artifacts.
• Data-driven gating (DDG) is a hardware-free alternative to traditional gating methods like belt gating (BG).
• Continuous-bed-motion (CBM) PET requires specialized DDG approaches for effective motion compensation.

Purpose of the Study:

• To investigate a CBM-capable DDG algorithm for whole-body PET/CT.
• To compare the performance of DDG against BG using optimally gated (OG) and elastic motion correction (EMOCO) reconstructions.
• To evaluate lesion metabolic volume, SUV measurements, and visual delineation quality.

Main Methods:

• Fifty-six patients with suspected thoracic or abdominal malignancies underwent 18F-FDG CBM PET/CT.
• DDG and BG signals were acquired and analyzed for correlation.
• Static, OG, and EMOCO reconstructions were performed for both gating methods.
• Quantitative (SUVmax, SUVmean, metabolic volume) and qualitative (lesion delineation) assessments were conducted.

Main Results:

• A moderate global correlation (0.48 ± 0.11) between BG and DDG signals was observed, with higher correlation in visceral regions (0.89 ± 0.07).
• All motion-compensated reconstructions (BG-OG, DDG-OG, BG-EMOCO, DDG-EMOCO) showed significantly higher SUV measurements than static reconstructions (P < 0.001).
• No significant differences in SUV or lesion delineation quality were found between BG and DDG methods within the same reconstruction type (OG or EMOCO).

Conclusions:

• DDG-based motion compensation in CBM PET acquisitions significantly enhances image quality compared to static reconstructions.
• The CBM-capable DDG algorithm provides performance comparable to traditional BG approaches.
• This DDG algorithm represents a valuable, hardware-free alternative for motion compensation in CBM PET systems.