SUV measurements from images reconstructed with the block sequential regularized expectation maximization algorithm: comparison of motion corrected vs. non-motion corrected data

3120 Objectives: PET-based deviceless gating (Q.Static®) was recently introduced as a software enhancement to the SiPM-based Discovery MI PET/CT scanner (GE Healthcare). In brief, it uses data-driven gating techniques capable of prospectively analyzing PET list data to detect motion and reconstuct images using data from the quiescent period only. We started using Q.Static® acquisition and reconstruction in June 2019, in addition to our standard block sequential regularized expectation maximization (Q.Clear®) reconstructions. The aim of the study is to determine if SUV measurements are significantly different between non-gated and gated reconstructions in standard organs. Methods: A preliminary analysis of 169 consecutive patients whose images were acquired with deviceless respiratory gating turned on yielded 17 patients whose detected motion in the areas of interest for this investigation met the system threshold required to trigger gated acquisition and reconstruction. All patients in this cohort were injected with 18F-FDG for standard of care clinical scans, with a mean injected dose of 10.42mCi +/- 1.07 mCi . Both gated and non-gated simultaneously acquired PET data were reconstructed using Q.Clear®and a penalty factor (beta value) of 400. Data was analyzed using MIM software. A 40mm region of interest (ROI) was placed over the normal liver and right lung, and a 20mm ROI was placed over the aortic arch. We retrospectively collected the SUVmax and SUVmean measurements from these ROIs. We used GraphPad® software to determine if there is a statistically significant difference between the SUV measurements of gated and non-gated reconstructions. Results: The mean SUV for each data set was calculated, and the mean difference between the non-gated and gated data was determined (Table 1). There is a slight yet statistically significant increase in the SUVmax measurements of liver using gated reconstruction while the SUVmean measurements show a slight decrease using deviceless respiratory gating. Lung tissue shows a slight yet statistically significant increase in both SUVmax and SUVmean using gated reconstruction. Aortic arch SUV showed a statistically significant increase in SUVmax, but no statistically significant difference in SUVmean. Conclusions: Further investigation and a larger cohort of patients is necessary to determine the significance of these findings for clinical applications, as well as performing a direct comparison of the SUV changes in lesions versus healthy tissue. This should be weighed against the ability to identify more lesions using Q.Static®. Table 1. Mean Difference (MD) in SUVmax and SUVmin in Gated and Non-gated Reconstructions