Estimation of Blood Perfusion and Metabolic Heat Generation of Lung Tumor During Cryosurgery

Abstract

Cryosurgery applies very cold temperature to freeze tumor cells. For accurate treatment, monitoring the temperature field during this process is a must. In this work, we estimated parameters such as metabolic heat generation and blood perfusion rate which are vital to determine the temperature field during cryosurgery. We applied the Quasi-Newton and the Gauss-Newton inverse algorithms to estimate these parameters. Temperature measurements were taken from sensor output. To estimate the parameters, the one-dimensional Pennes’ bioheat equation was utilized. The parameters were predicted using least square minimization of the difference between sensor output and estimated temperature values. Once the parameters are obtained, the temperature distribution around the lung tumor at any time can easily be determined. Random initial values were given for both algorithms. The result showed that the Gauss-Newton method has faster convergence rate as compared to the Quasi Newton method in estimating the target parameters. The output of the research will help cryosurgeons to monitor the temperature of the cryoprobe during cryosurgery procedures.