Augmented reality-enhanced endoscopic images for annuloplasty ring sizing

Abstract

Mitral valve annuloplasty is done in patients with mitral valve insufficiency in order to stabilize, remodel or downsize the often symmetrical or asymmetrical dilated annulus by stitching a prosthetic ring on this anatomical structure. Prosthetic rings are available in different shapes and sizes. State-of-the-art intraoperative sizing techniques for determination of the appropriate ring prosthesis are ambiguous and highly depend on the surgeon’s expertise. We propose a new augmented reality environment for visualizing the prosthetic ring in-situ on endoscopic images and therefore aid in ring selection. The superimposed ring gives quantitative information and visual cues allowing to compare the selected ring prosthesis with the patient’s annulus. Furthermore, it helps in determination of regions where an asymmetrical dilatation can be observed. Our method identifies 2D points on the endoscopic images by detecting the entry points of the mattress sutures into the annular tissue for ring fixation. 3D shape information of the annulus are obtained from ultrasound images of the patient. The pose estimation problem of the 3D annulus model is solved using an adapted iterative closest point algorithm. Neither additional hardware nor placement of artificial fiducial markers are required by the proposed approach.