(A) rigid motion compensation using covariance matrix adaptation evolution strategy(CMA-ES) for Cone-Beam CT

Abstract

As the importance of preventive medicine is emerging, early detection of disease through computed tomography is becoming important. However, since radiation is harmful to human body, it is better to avoid meaningless exposure. One of such nonsensical exposure types is artificial artifact that occurs due to the movement of the subject when exposure. Various studies have been carried out to solve this artificial artifact. Recently, researches using an evolutionary algorithm have been spotlighted. In this paper, I propose a method to compensate the translation movement of the object using the Covariance Matrix Adaptation Evolution Strategy. The proposed method compensates the motion of the object in the direction of optimizing the cost function composed of the gradient variance and the euclidean distance of the parameters using the covariance matrix adaptive evolutionary strategy. The effectiveness of the algorithm for various motions is verified through simulations.