A Heat Conduction Simulator to Estimate Lung Temperature Distribution During Percutaneous Transthoracic Cryoablation for Lung Cancer

Abstract

To develop an evaluation method for the curative field when using X-ray CT imaging during percutaneous transthoracic cryoablation for lung cancer, we constructed a finite-element heat conduction simulator to estimate temperature distribution in the lung during cryo-treatment. We calculated temperature distribution using a simple two-dimensional finite element model, although the actual temperature distribution spreads in three dimensions. Temperature time-histories were measured within 10 minutes using experimental ex vivo and in vivo lung cryoablation conditions. We adjusted specific heat and thermal conductivity in the heat conduction calculation and compared them with measured temperature time-histories ex vivo. Adjusted lung specific heat was 3.7 J/(g·°C) for unfrozen lung and 1.8 J/(g·°C) for frozen lung. Adjusted lung thermal conductivity in our finite element model fitted proportionally to the exponential function of lung density. We considered the heat input by blood flow circulation and metabolic heat when we calculated the temperature time-histories during in vivo cryoablation of the lung. We assumed that the blood flow varies in inverse proportion to the change in blood viscosity up to the maximum blood flow predicted from cardiac output. Metabolic heat was set as heat generation in the calculation. The measured temperature time-histories of in vivo cryoablation were then estimated with an accuracy of ±3°C when calculated based on this assumption. Therefore, we successfully constructed a two-dimensional heat conduction simulator that is capable of estimating temperature distribution in the lung at the time of first freezing during cryoablation.