UNMET NEED
The standard techniques for many orthopedic surgical operations often become a high risk, expensive procedure that rely on the ability of a surgeon to mentally reconstruct computed tomography (CT) images and co-register these images with the patient’s 3-D anatomy intraoperatively. This typically requires multiple X-ray images at several orientations of the CT mounted C-arm; subjecting the patient to significant radiation exposure, extended time under anesthesia, and increased operating room costs.
TECHNICAL DESCRIPTION
3-D augmented reality technology allows intuitive real-time overlap of 3-D Cone-Beam CT (CBCT) reconstruction volumes and visualizations of the patient’s body surface. This invention uses minimally modified off-the-shelf imaging equipment (CBCT imaging systems and 3-D C-arms), a mounted RGB-D depth camera, and custom image registration/stitching software. The software uses machine learning algorithms to reconstruct CT images, segment the patient’s body surface, co-register the CT volumes with the patient’s surface, and provide a display overlaying the images to guide surgical intervention.
STATUS OF DEVELOPMENT
A prototype augmented reality system was created using a Siemens Arcadis Orbic 3-D C-arm, an Intel RealSense RGB-D camera, and a medical grade PC. A usability study was conducted over 21 procedures performed by surgeons on phantoms to compare the augmented reality prototype against standard methods. The study tested the duration, quantity of X-ray images acquired, placement accuracy, and surgical task load; finding a significant improvement across all metrics using the augmented reality prototype.