Genetic optimization methods for medical image registration

We solved the problem of homogenous landmarks in the procedure of medical image registration via deformable contour model, and genetic algorithm (GA) was used to search the optimal results. The GA used the energy function of the deformable contour as its fitness function, and an adaptive search space scaling technique was presented which was operated by dynamic mapping. The experimental results showed that the proposed method could efficiently establish the correspondence between medical atlas and medical images, which was an important improvement to the previous interactive methods.