Guidewire tracking with projected thickness estimation

Real-time guidewire tracking provides valuable navigational aid during endovascular procedures and is an important problem in computer-aided interventions. We propose an extension to the background estimation tracking method in order to enable estimation of the projected guidewire thickness. First, we use a combination of MAP estimation and Kalman filtering for background modeling. Second, the obtained background is used for guidewire tracking and for projected thickness estimation. The x-ray data acquisition model is used to estimate the thickness. The projected guidewire thickness provides insight into the 3D position of the object and can be used to improve monoplane 3D tracking. Estimated position of the guidewire compares favorably to the existing state of the art methods such as Hessian eigenvalue analysis.