Medical Image Registration Based on Mutual Information and Hybrid Algorithm
Lv Yaoming, Jiang Dayu, Yuan Tian, Xiao Peiyan, Qing Chang, Shi Jin-chi · 2021
Aiming at the problem that the existing traditional medical image registration algorithm is easy to fall into the local extremum and low registration accuracy, this paper proposes a hybrid optimization algorithm which combines the downhill simplex algorithm and firefly algorithm as an optimization strategy and takes mutual information as the similarity measure to complete the chest image registration tasks. The hybrid optimization algorithm takes the parameter optimization result of decreasing simplex as the initial solution of the firefly algorithm, and further optimizes to obtain the global optimal value. In the case of local optimum, it can avoid the mutual information function when the downhill simplex algorithm is used alone., and it can effectively improve the computational efficiency of the firefly algorithm. The experimental results show that the proposed algorithm can improve the registration time and registration effect significantly, and can complete the medical image registration tasks better.