3D visual stabilization for robotic-assisted beating heart surgery using a thin-plate spline deformable model

Visual stabilization in robotic-assisted minimally invasive beating heart surgery is used for avoiding rhythmic motion of the heart. A robot, which is remotely controlled by the surgeon, can automatically compensate the motion of the beating heart and a stabilized image of the heart is shown to the surgeon. The visual stabilization cancels the undesirable movements of the heart while preserving variation of colors and textures for example cuts on the heart surface. Unlike the existing methods that used landmarks, the proposed approach uses a 3D region based tracking algorithm for visual stabilization. In this paper, a modified version of 3D tracking algorithm using a thin-plate spline deformable model is developed to improve the motion tracking. Afterwards, the 3D image stabilization based on bilinear interpolation is used for motion compensation. The performance of the proposed approach is evaluated in recorded stereo image sequence of an in-vivo using DaVinci surgical robotic platform. Experimental result demonstrates high quality of the visual stabilization. A novel approach developed in this paper which we call it adaptive reference change approach.