Respiratory Data Driven Gating Workflow in Clinical PET Imaging -a Comparison to External Driven Gating

Introduction: Respiratory motion (RM) in PET imaging can affect both qualitative and quantitative evaluation of lesions, leading to inaccurate uptake, increased tumor volume and reduced contrast. RM correction based on the quiescent phase of the respiratory cycle (RC) has already successfully been used clinically, though traditionally it relies upon externally driven gating (EDG). Such gating option although relatively accurate, substantially affects clinical workflow in terms of patient preparation, scan setup and data acquisition duration. In this work, using the first clinically available data driven gating (DDG) software, we report on its evaluation from a workflow perspective and compare findings against EDG. A total of 165 patient datasets using a variety of clinical indications and tracers, were prospectively acquired on an SIPM based 25cm FOV PETCT system (GE Discovery MI, GE Healthcare, USA) of which 154 using DDG and 11 using EDG (RPM, Varian Medical Systems, USA). The difference between the 2 gating methods lies on the decision mechanism used to select the bed for MC using the quiescent phase of the RC (Q.Static). In Q.Static mode, bed acquisition time could be extended to create images out of an equivalent amount of events as when using data from the whole RC. For this study, scan time was doubled for beds selected for motion screening. While for the EDG setup a fix number of beds was used (2 beds covering lungs/liver/upper abdomen), for prospective DDG the software, using solely the raw PET data and a pre-selected user-defined motion screen threshold (default R=15), automatically takes an on-the-fly bed-by-bed decision whether motion has been detected, extends the acquisition as prescribed and corrects for motion. Workflow evaluation between the 2 gating options is analyzed both in terms of pre/post-acquisition patient and protocol setup as well as scan time changes during data acquisition. In terms of pre/post acquisition patient and protocol setup, using DDG resulted in an average time reduction of 4min 10sec compared to EDG for MC. The largest changes were observed with respect to positioning and verifying the tracking tool on the patient's chest/abdomen, setting up the prescribed protocol and setting up and saving the waveforms on the EDG computer. With respect to scan time changes during acquisition, the benefit of DDG directly depends on the predefined motion threshold selected. Figure 1a shows the total number of beds in which DDG is be