Stochastic and exponential analysis of precordial washout curves for myocardial blood flow measurement. Experimental evaluation.

Publisher Website
Google Scholar
Abstract
133Xe washout curves were recorded from the anterior wall of the left ventricle down to at least 1% of the peak following bolus injection into the bypassed left anterior descending coronary artery of dogs. Washout curves were obtained for a wide range of independently measured flows (0.36-9.57 ml/min per g). Curves (36) were analyzed to evaluate the effect of tracer recirculation and flow inhomogeneity (assessed by simultaneous microsphere injection) on the flow values derived from monoexponential fitting of the curve and from stochastic analysis. The results were compared to those obtained by computer model simulation. Although minimized, recirculation significantly affected the evaluation of flow by both methods of analysis (monoexponential fitting down to 30% of the peak underestimated actual flow by an average of 15%, stochastic analysis by 45%). Correction for recircualtion by the dual injection method led to correct estimation of flow by monoexponential fitting; stochastic analysis produced an average underestimation of 24%. The failure of stochastic analysis to estimate flow correctly was explained by a lower observed peak activity caused by noninstantaneous input. As predicted on the basis of a multicompartmental model, transmural flow inhomogeneity as great as that explored (a 3-fold difference in flow between the best and worst perfused thirds of the wall) did not significantly affect the computation of flow by monoexponential analysis. Flow-limited, Xe blood-tissue diffusion cannot be supported by this study, as the underestimation of actual flow observed at high flow rates can be ascribed to the effect of tracer recirculation.

This publication has 24 references indexed in Scilit: