Quantification of regional myocardial blood flow using 13N-ammonia and reoriented dynamic positron emission tomographic imaging.

BACKGROUND Regional myocardial blood flow has been quantified using transaxial positron emission tomographic (PET) imaging and tracer kinetic modeling. However, the use of transaxial images limits the accuracy of regional partial volume corrections and the localization of the quantified regional flow values. The purpose of the present study was to overcome both problems by calculating regional flows from reoriented short-axis PET images. METHODS AND RESULTS Twelve experiments were performed in four dogs. 13N-ammonia was injected intravenously while microspheres were administered into the left atrium during baseline, hyperemic, and low-flow conditions. Serial transaxial frames were acquired with a 15-plane PET scanner and reoriented into short-axis frames. The arterial input function and eight regional myocardial tissue activity curves were derived from the reoriented frames. The arterial input functions were corrected for ammonia metabolites, and the myocardial tissue curves were corrected for spillover of activity, partial volume effects, and heterogeneities in the image's spatial resolution introduced during reorientation. Corrections for regional partial volume were based on estimates of the regional myocardial activity thickness derived from reoriented diastolic images of the heart. The myocardial 13N-ammonia kinetics were described with a two-pool compartmental model. Values of regional myocardial blood flow by PET correlated linearly with those by microspheres (slope, 0.94; y intercept, 0.06 ml/min/g; r = 0.93) over a wide range of flows. CONCLUSIONS Regional myocardial blood flow can be measured accurately and noninvasively from serially acquired and reoriented short-axis 13N-ammonia images, thus overcoming limitations inherent to the use of transaxially acquired images and permitting a more complete evaluation of regional blood flows throughout the left ventricular myocardium.