Markerless Navigation

The ability to register the bone’s anatomy to a preoperative plan and track the patient’s movements intraoperatively is crucial for any form of Computer-Assisted Orthopaedic Surgery (CAOS). Currently, this requires the use of dedicated optical markers, and manual collection of landmark points using tracked instruments, increasing the length and invasiveness of the procedure. Our group has conducted pioneering work on the use of modern 3D stereo camera technology coupled with Artificial Intelligence, in order to automatically detect and segment the patient anatomy and track it intraoperatively without the need of external markers [1]. We have also developed intraoperative assistive planning tools that help the operating surgeon select an optimal prosthesis for a given patient and optimise its planned positioning. 

Augmented Reality

 

We have also explored using immersive methods for human-computer interaction during robot-assisted surgeries including the use of augmented reality during knee arthroplasty. In our group, we have developed robust, user-agnostic techniques for co-registering and positioning virtual models and objects in directly on patient anatomy in meaningful locations [2]. Work conducted in this project has developed a fully AR-centric workflow for joint-replacement procedures with some preliminary pilot studies done to examine the impact of using AR on relevant user performance metrics [3].