A computer implemented method and a system for registering and tracking an anatomical structure of a patient. A method includes acquiring a first stereoscopic image dataset including video images having different viewing angles, and providing a first image content showing the anatomical structure; registering the first image dataset to a 3D image dataset showing the anatomical structure; acquiring a second stereoscopic image dataset comprising video images having different viewing angles and providing a second image content showing the anatomical structure, which differs from the first image content; and tracking the anatomical structure based on the second image dataset.