333. A personalized acquisition protocol to optimize 18F-FDG PET/CT whole-body imaging

Abstract

Purpose PET image quality degradation is encountered in over-weight and obese patients even when administered activity is increased with body weight. For these cases is more appropriate to increase the acquisition time rather than administered activity. Aim of the work is to evaluate a metric related to patient body constitution to optimize 18F-FDG PET/CT Whole-Body (PET-WB) acquisition procedure to guarantee accurate diagnosis with minimal radiation exposure of patients. Methods PET-WB acquisitions were performed with the TruePoint Biograph scanner (Siemens) with 216 mm axial FOV; images were reconstructed with OSEM-3D with 2iterations/8subsets. Initial retrospective analysis was performed on 79 patients studied according to the PET-WB routine protocol: injection of 4.0 ± 0.6 MBq/kg, with a maximum of 500 MBq in obese subjects; acquisition duration of 2 min/bed and 3 min/bed respectively for body weight ⩾ 100 kg. Signal to noise ratio (SNR) measured on liver region was taken as representative parameter of image quality. Diabetic patients and those with hepatic lesions were not included in the analysis. Measured SNR values were normalized to injected activity and acquisition duration (SNRn) and correlated with patient-specific parameters: height to weight ratio (HWR), body mass index (BMI), body surface area (BSA), lean body mass (LBM), fat free mass (FFM). A mathematical model was extrapolated from the parameter most correlated to SNRn values. The developed model was validated prospectively on others 66 patients. Results Retrospective analysis showed a significant SNR decrease with increasing patient size. SNRn values fitted to patient-specific parameters showed the highest R2 (0.86) with HWR. Setting SNR reference value of 9, derived from an image quality deemed acceptable by clinicians for normal-weight patients, the HWR mathematical model allowed to personalize acquisition duration and keep constant image quality for different weight class. By fixing administered maximum activity to 380 MBq for obese subjects an increase of 30% acquisition time corresponded to 30% image SNR gain, with a 30% dose reduction to patient with respect EANM recommendations. Conclusions The application of a model based on specific patient HWR allowed to improve SNR by personalizing acquisition duration without increasing patient dose exposure and to standardize PET-WB procedure.