Multi-modality registration by using mutual information with honey bee mating optimization (HBMO)

Registration is a popular technique commonly used in medical image processing. In this paper, we propose a new registration algorithm which uses the mutual information (MI) as the similarity measurement function and utilizes a new optimization algorithm, called honey bee mating optimization (HBMO), to obtain the optimal registration. By simulating the biological evolution, HBMO can obtain a set of optimized parameters for registration with the largest similarity measure. By applying the proposed method to medical images, the experimental results showed that the method achieved better accuracy in registration than the conventional Powell's optimization method which is the most commonly used method in medical image registration. Also, the proposed method remained stable and accurate during the experiment of using several different source images. With each source image, we calculated mean ± SD of the parameters by repeating twenty times.