Data-driven gating (DDG)-based motion match for improved CTAC registration

Area of Science:

• Medical Imaging
• Radiology
• Nuclear Medicine

Background:

• Respiratory motion causes artefacts in thoracic PET/CT imaging, leading to inaccurate attenuation correction.
• Misregistration between PET and CT images due to arbitrary helical CT acquisition phases results in errors in tumour delineation and lesion uptake values.
• Motion artefacts in PET/CT imaging can be minimized using a data-driven gating (DDG)-based motion match (MM) algorithm.

Purpose of the Study:

• To develop and evaluate a DDG-based MM algorithm for phase-matching CT images to a specific respiratory phase (end-of-expiration).
• To assess the algorithm's effectiveness in reducing motion artefacts in PET/CT imaging.
• To compare algorithm-generated motion-matched CT (MMCT) images with ground truth four-dimensional CT (4DCT) phases.

Main Methods:

• A DDG-based MM algorithm was developed to estimate the respiratory phase of helical CT (CT) data.
• The algorithm warps the CT data to a specified respiratory phase, enabling phase-matching with PET data.
• Four-dimensional CT (4DCT) data acquired concurrently with PET/CT served as the ground truth for comparison; liver and lesion margin positions were measured.

Main Results:

• No significant differences were observed in lesion measurements between 4DCT and MMCT.
• A minor but significant difference was noted in liver measurements between 4DCT and MMCT.
• The MM algorithm successfully generated phase-matched CT images equivalent to 4DCT, demonstrating its efficacy.

Conclusions:

• The MM algorithm effectively phase-matches helical CT to the end-of-expiration phase of a ground truth 4DCT.
• This method reduces motion artefacts in PET/CT imaging caused by registration issues.
• The algorithm achieves motion correction without necessitating additional radiation exposure for the patient.