Evolutionary computation in identification of a tumor

The paper deals with an application of evolutionary computation in identification of shape and position of a tumor region in the biological tissue domain. The problem is formulated as an inverse problem that is solved by the minimization of a functional formulated as some distance between measured and computed skin surface temperatures. The functional is considered as the fitness function and the minimization is performed by an evolutionary algorithm with the floating point representation of chromosomes. Geometrical parameters or shape and position of the tumor play the role of genes. The evaluation of the fitness function is preceded by the solution of the direct problem for the bioheat transfer (the Pennes equation) by means or the finite element method. Numerical examples of evolutionary computation for 2-D problems are presented.