Modelo de Erro de Aproximação para estimativa de fluxo de calor na Equação de Biotransferência de Calor Unidimensional

Abstract

This work deals with the application of the approximation error model in the estimation of the boundary heat flux function in an inverse bioheat transfer problem. In general, complete models involving bioheat transfer have a high computational cost. The objective of this work is applying a reduced model, based on the classical heat conduction equation, together with Bayesian inference techniques in order to estimate parameters in a biological problem. From simulated measurements, generated by the solution of the complete model, the estimates in the inverse problem are obtained by applying the Monte Carlo Method with Markov Chain involving a reduced model and an approximation error model. The results reveals that the technique used was capable to produce accurate results, allowing the estimation of the heat flux in a biological problem, using a small computational time model, which in this case was a classical heat conduction equation.