Design and thermal modeling of a non-invasive probe for measuring perfusion by thermodiffusion

Abstract

This research 1) explores the feasibility of developing a non-invasive probe to precisely quantify microcirculatory blood flow (tissue perfusion), in real time and in absolute units, and 2) presents designs and models of such a probe, along with an evaluation of various design-model combinations. Bowman et al. have developed an invasive thermodiffusion probe that measures tissue perfusion accurately, continuously, and in real time. This method employs a self-heated thermistor placed in perfused tissue. From a knowledge of the power required to heat the thermistor probe to a given temperature, perfusion can be calculated using an analytical or numerical model. Using Bowman's thermodiffusion probe (designed for invasive use) in a non-invasive manner, a perfusion study was performed. The data clearly show the promise of a non-invasive thermodiffusion perfusion probe (designed for non-invasive use), and the design of such a probe was pursued by adapting the invasive technology for a non-invasive probe. Because perfusion is not actually measured but calculated from measured quantities by a model of the probe and perfused tissue, the design of the non-invasive probe occurred hand-in-hand with the development of analytical models.

(cont.) The results of the clinical study are presented, as well as two designs together with possible one-dimensional analytical models. Using a finite-difference model of the two probe designs and the underlying perfused tissue, the errors that result from approximating these designs as one-dimensional models have been determined. It is shown that modeling a thin, disk-shaped thermistor probe as a hemisphere of appropriate radius can result in an error in calculated perfusion which is small enough for clinical use.