Perils and Pitfalls in Quantitative Pet Phantom Measurements.

2512 Objectives This work reports both strategies for accurate and reproducible PET phantom measurements, and also identifies several subtle, but correctable pitfalls associated with the filling and imaging of common phantom platforms. Methods Three common PET phantoms (NEMA IQ phantom, 20cm uniformity phantom, and CTN oncology phantom) were routinely imaged on four PET/CT scanner systems to assess scanner performance. However, several unexpected pitfalls were encountered. 1) Some of the FDG in the aqueous solution in the phantom was metabolized by the microorganisms that were living on the surface of the phantom walls. A detailed sanitization with a safe bleach dilution is now routinely performed. 2) The PET scanner’s nonlinear response to total activity in the FOV unexpectedly impacted quantitation. Consequently, the scanner’s count rate response to varying activity was mapped and an optimal, no-defect, activity range was identified. 3) PET/CT scanner’s inaccuracy in the assessment of attenuation correction of epoxy and other phantom materials was investigated. 4) Ga-68 alone and its solution forms with EDTA and DOTATOC were examined for their binding with acrylic phantom walls. Results After phantom sanitization, quantitative errors reaching as high as 12% were resolved to less than 2%. Careful cross calibration and regular quantitative performance monitoring resulted in consistent results (within ±2%) over time. Dose calibrator geometry effects were reported to be up to 5% for differing syringe volume, so constant volumes are now used. Scanner count rate effect was on the order of 1% per mCi. Conclusions PET/CT scanners when appropriately cared for are remarkably stable quantitatively accurate measurement devices. Phantom studies, when carefully performed can not only characterize the quantitative performance, but serve as an early warning to problems in accuracy and consistency when arising.