Tracer Kinetic Modeling by Morales-Smith Hypothesis in Hepatic Perfusion CT

Abstract

Most of the existing tracer kinetic models for dynamic contrast-enhanced CT or MRI do not fully describe the principles of intra- and transcapillary transport of tracers. One point is to disregard the concentration profiles between the inlets and outlets of capillaries, which may cause a biased estimation of tissue parameters by a systematic error. The Morales-Smith hypothesis enables one to resolve this ambiguity by assuming that the difference between arterial and venous concentrations is proportional to the difference between the arterial and capillary concentrations. If the backflow of administered tracer into the plasma compartment is negligible compared to its outflow into the interstitial compartment during the initial enhancement phase after tracer administration, the capillary concentration can be considered to fall exponentially along the capillary from the arterial concentration to the venous concentration by the Renkin-Crone model, i.e., unidirectional extraction fraction, which can be incorporated in the concept of the Morales-Smith hypothesis. In this study, we reformed the mass-balance equations and mathematical solutions of several representative and well-known tracer kinetic models so that the Morales-Smith hypothesis could be incorporated into their compartment tracer kinetics, considering a tissue-specific factor independent of time as proposed by Brix et al. [5]. The tissue-specific factor was applied to a liver tumor case study in perfusion CT to illustrate the potential effectiveness of the Morales-Smith hypothesis. The proposed scheme was shown to be potentially useful for more consistent and reliable estimation of physiologic tissue parameters.