TY - JOUR T1 - Myocardial flow estimation using a truncated 82Rb PET acquisition JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 203 LP - 203 VL - 52 IS - supplement 1 AU - Jeffrey Kolthammer AU - Raymond Muzic Y1 - 2011/05/01 UR - http://jnm.snmjournals.org/content/52/supplement_1/203.abstract N2 - 203 Objectives We examined the accuracy and precision of flow estimates as a function of scan duration in a 82Rb PET myocardial exam. Our hypothesis was that flow information is concentrated in the time period immediately following activity administration so a short-duration scan could be sufficient. Methods Typical time activity curves (TACs) for an image-derived input function and tissue region were generated from experimental data. A one-compartment kinetic model was used to estimate an index of flow K1, backward constant k2 and blood-tissue fraction fv. The effect of noise on estimated parameter values was analyzed using Monte Carlo simulation with simulated noise ranging from 1% to 50% standard deviation (SD). Also, based on an assessment of the experimental design using the D-optimality criterion, scan durations ranging from 0.8 to 4.0 minutes were considered. For each noise level and scan duration, parameter values were estimated for 1000 realizations. Estimation error was calculated as a percentage of the true parameter value. Estimation errors were visualized via a box-plot and bias and precision quantified using mean and SD. Results Estimation error increased with noise level. With the longest scan duration, 4 min, precision of K1, k2 and fv estimates were, respectively, 25%, 65% and 93% SD with 50% simulated noise and 14%, 39% and 22% with noise typical of clinical studies. Reducing the scan duration from 4 to 2.5 min caused only a slight worsening of K1 precision, with SD increasing from 14% to 19%; further reducing to 1.5 min increased SD to 40%. k2 required a longer scan duration: SD increased from 39% (4 min) to 87% (2.5 min). Bias in K1 was observed (p<0.01) and was greater than 10% for durations shorter than 1.5 min. Conclusions PET acquisitions used only for estimating flow from the K1 parameter of a one-compartment model may be shortened significantly from those typically used for MPI, while retaining much of the available precision. Research Support State of Ohio Department of Development: Ohio Third Frontier Grant TECH-11-04 ER -