Estimation of myocardial perfusion using deuterium nuclear magnetic resonance

The studies described in this dissertation were designed to evaluate the utility and validity of measurement of myocardial perfusion by deuterium NMR spectrocopy of deuterium oxide (D$\sb2$O) washout. The D$\sb2$O washout method of perfusion measurement has several advantages over other methods: the method does not require removal of the heart; the tracer is not radioactive and has been safely used for other clinical diagnostic purposes; the method can be combined with NMR spectroscopy of other nuclei and with other methods for measurement of the heart\u27s mechanical function and metabolic state. Washout of tracer from the myocardium follows a decaying exponential whose rate constant is the reciprocal of the perfusion rate. This interpretation of tracer washout as a decaying exponential is derived from the Fick principle. It is also shown that neither the natural abundance of deuterium in body water nor residual tracer from previous administrations affect the measurement. Exponential interpretation of the Kety-Schmidt equation assumes that the concentration of tracer in the arteries supplying the organ of interest is much smaller than the concentration in the organ at the time of measurement. This assumption was verified in the experimental model. Laminar flow of blood within the vessels between the point of tracer injection and the point of measurement causes the tracer input function to have long duration, which causes underestimation of perfusion rate. It also causes observed tracer washout curves to appear biexponential. Local inhomogeneity of perfusion is another cause of the apparent biexponentiality observed in this and other tracer washout studies. The method was validated by comparison to simultaneous volumetric measurements of flow in the isolated perfused rat heart. Flow was measured by diverting the perfusion output to a graduated cylinder. In vivo validation was by comparison to results of a Doppler technique. In both cases, myocardial perfusion rates obtained using the D$\sb2$O washout method agreed with the volumetric results. Perfusion values obtained with this method also agreed with previously reported values obtained using other tracers and/or radiolabeled microspheres