non-vulnerable adults undergoing spine surgery, with a posterior approach

Spine surgery, such as fusion requiring the need for precise intraoperative registration and navigation

Success in recording 25 RGB-D videos of spine surgeries along with corresponding CT scans

Collected samples should contain a wide variety of exposed vertebrae, including cervical, thoracic and lumbar vertebrae

Patients undergoing surgery should be diverse in terms of ethnicity, diversity being the % of patients who are not 'white british'

Patients undergoing surgery should be diverse in terms of age, diversity being captured by range and standard deviation measures

Outcome is to leverage the recordings to train a markerless tracking system that will intraoperatively track the pose and location of the spine within the surgical site. The trained system will be benchmarked against a test subset of the collected videos. The success of the trained system will be determined by its accuracy in automatically tracking the pose of vertebrae features within the test subset videos, against the labelled pose indicated in the video by the operating surgeon.