Abstract
2003
Objectives: Positron emission tomography/computed tomography (PET/CT) imaging is established as essential for cancer diagnosis and treatment monitoring. However, PET/CT imaging of the lung and abdomen can be compromised by respiratory motion. To address this, several respiratory gating methods have been developed that use device-based respiratory trigger signals as well as predefined protocols. Experience has shown that these methods can be cumbersome and often provide little or no benefit. Therefore, being able to achieve higher sensitivity PET/CT imaging without the need for external triggering devices or custom protocols would facilitate routine PET/CT acquisitions with respiratory motion gating.
Methods: We used a GE Discovery MI (DMI) PET/CT scanner with a 25 cm axial field-of-view and NEMA sensitivity of 21cps/kBq [1]. Respiratory gating was performed with a data-driven gating (DDG) algorithm that does not require an external triggering or gating device [2]. The DMI+DDG combination was used in the scanning of 122 patients at the Seattle Cancer Care Alliance in September through November 2018. Images were produced using full respiratory gating and quiescent-period gating [3]. PET images were analyzed for lesion uptake, metabolic volumes, respiratory shifts of lesions, and diagnostic image quality. We also evaluated the impact on workflow from a technologist’s perspective.
Results: Out of the 122 procedures carried out with respiratory gating, approximately 87% of the PET/CT procedures were performed with the existing standard of care protocol incorporating DDG. These procedures did not increase time to the acquisition. In 13% of the cases, existing standard of care protocols added extra scanning time, prescribed prospectively, in anatomical areas affected by respiratory motion in order to compensate for loss of statistics when utilizing quiescent period motion correction. The use of DDG yielded an average of 14% increase in SUVmax and 17% reduction in lesion volume when compared to the non-gated images. In some cases however, the changes in SUVmax and lesion volume were much larger. The results collected from the workflow survey indicated that the user interface for setup of prospective and retrospective respiratory-gated PET/CT acquisitions was easy to understand. The additional time used in the scan setup process to enable DDG was minimal, and the protocols allowed the prospective choice of which bed position received motion correction and/or extended time for acquisition. Previously acquired list-mode data acquisitions can be assessed retrospectively with DDG and corrected for respiratory motion if desired.
Conclusions: The use of the data-driven gating (DDG) algorithm with dual list-mode and histogram acquisitions allowed for routine respiratory gating in clinical whole-body PET/CT acquisitions without impacting workflow or scan duration. Images can be produced using full respiratory gating and quiescent-period gating. The cases evaluated revealed multiple instances of lesions with reduced metabolic volume and increased maximum SUV, when compared to the non-gated images.