Determining optimal F18-FDG dose in clinical whole body PET acquisitions using Acquisition-Specific NEC

Objectives: The Acquisition Specific Noise Equivalent Count (AS-NEC) method [1,2] takes data from a clinical patient study and computes the relative NECs which would have been obtained if a different amount of activity had been injected into that patient. The point at which the AS-NEC curve is maximized corresponds to the optimal, in a signal-to-noise sense, amount of injected activity for imaging that frame of that patient. We apply AS-NEC analysis FDG oncology studies to determine the optimal injection dose for PET scans using the Discovery IQ (DIQ) 5 ring PET/CT scanner. Methods: The dead time correlation factors required to compute the AS-NEC formulation were determined by analyzing decay series data from a NEMA count rate analysis phantom (70 cm line source in a 70 cm long, 20 cm diameter solid phantom). The factors were validated using data from a separate decay NEMA decay series and a decay series using a 20 cm long, 20 cm diameter flood phantom. The AS-NEC model was then applied to data from 50 clinical FDG PET scans, representing a patient population with a wide range of weights (41-180 kg) and activity injected (66-600 MBq). Individual frames that included the bladder or brain were excluded from the analysis. The injected activity was decay-corrected to the scan start time and then the AS-NEC results were applied to determine the optimal dose, decay-corrected to scan time, for each frame of each patient study. Results: The accompanying figure shows the patient activity corresponding to peak AS-NEC for each acquisition frame as a function of patient weight. Linear regression indicates that the optimal activity at the start of scan is given by 203+0.24[asterisk](weight in kg). This corresponds to an injection of roughly 300+0.35[asterisk](weight) MBq for a typical 60-minute delay from injection to scan, and about 325 MBq (8.8 mCi) for a nominal 70 kg patient. Conclusion: We have calibrated the AS-NEC technique for use on data from the Discovery IQ PET/CT scanner, and used that technique 50 FDG oncology studies to determine the dose to achieve optimal image SNR in those images. The results indicate that there is an advantage to increasing the dose somewhat in heavier patients, but not by as much as a dose-proportional-to-weight regime would indicate.