TY - JOUR T1 - Minimization of PET scanner calibration variability for multi-center clinical trials utilizing NIST traceable sources JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 202 LP - 202 VL - 52 IS - supplement 1 AU - John Sunderland AU - Levent Sensoy AU - Justin Otis Y1 - 2011/05/01 UR - http://jnm.snmjournals.org/content/52/supplement_1/202.abstract N2 - 202 Objectives Quantitative PET scanner calibration is both a critical and controllable factor when PET imaging is used in multi-center clinical trials. Published literature suggests >10% variability when a single phantom is imaged at multiple sites. Careful calibration procedures using positron-emitting NIST traceable sources should reduce variability in a multi-center clinical trial setting. Methods To simulate a multi-center environment, 3 PET/CT systems were used, each calibrated using a different commercial Ge-68 20 cm uniform phantom. Six different dose calibrators (DC) were used, each with DC gains individually set to match the activity of a commercial Ge-68 NIST traceable source. In each experiment an F-18 sample was assayed using each of the 6 DC. The F-18 activity was injected into a single 20 cm uniform phantom and imaged on the 3 PET/CT systems. 40 measurements for each scanner were performed over 3 months. Image-based activity concentrations for each scanner were compared with calculated activities based upon the 6 DC measurements and phantom volume. The phantom volume was measured by weight (±1g) and scanner time was matched to DC time (± 30 sec). As DC geometry effects were observable, a uniform syringe volume (1±0.1mL) was used. Primary inter- scanner accuracy and variability errors were revealed to originate from inaccurately assayed commercial Ge-68 calibration phantoms. For the final 13 (of 40) measurements, the PET/CT systems were recalibrated to match NIST traceable measurements by correcting Ge-68 phantom concentrations. Results When using NIST traceable DC measurements, meticulous experimental technique, and recalibrated Ge-68 phantom concentrations, scanner derived concentration measurements agreed well with dose calibrator based measurements. All scanner/DC combinations measurements agreed (±1%) over the final 13 measurements taken over 2 months. Conclusions Tying PET scanner and DC measurements to a NIST traceable positron-emitting source significantly reduces measurement errors within a multi-center clinical trial setting. Research Support NIH U01 CA14020 ER -