Influence of respiratory motion on lesion quantification by FDG PET/CT

1170 Objectives To assess the influence of respiratory motion on the activity and volume quantification of a radioactive lesion and the value of respiratory gating to correct artefacts induced by this motion. Methods All acquisitions were performed on a Gemini XL PET/CT (Philips Medical Systems). First we quantified the influence of respiratory motion on a NEMA quality image phantom (containing 6 spheres of different diameters) mounted on a mobile platform conceived in our department to simulate the respiratory motion in Y-Z directions. PET/CT images of the phantom with and without motion were acquired. FDG PET/CT whole-body exam followed by a respiratory gated acquisition using a respiratory bellows belt was performed for 7 patients. We used the Oasis v1.4.4 software (Segami Corporation) to compare, for different threshold segmentation methods, SUVmax and SUVmean, the metabolic volume and the total lesion glycolysis in the radioactive spheres, for the motion-free and the motion-induced phantom study and for the gated and non-gated PET/CT acquisitions in the patient study. Intra and inter observer reproducibility of measures were calculated. Results With the NEMA phantom, motion was shown to be responsible for a significant decrease of SUVmax (7.8 versus 11.3, p 86% for glycolytic volume. Conclusions Respiratory motion is responsible of a significant decrease of SUVmax and an increase of hypermetabolic volume. Our work confirmed the great robustness of SUVmax for the semi-quantitative analysis, making it reliable for therapeutic monitoring