Luke Haliburton1, Hooman Esfandiari2, Pierre Guy3, Carolyn Anglin4,5,6, Antony Hodgson7. 1. Department of Mechanical Engineering, University of British Columbia, Vancouver, BC, Canada. luke.haliburton@ifi.lmu.de. 2. Surgical Technologies Laboratory, Biomedical Engineering, University of British Columbia, Vancouver, BC, Canada. 3. Department of Orthopaedics, University of British Columbia, Vancouver, BC, Canada. 4. Department of Civil Engineering, University of Calgary, Calgary, AB, Canada. 5. Biomedical Engineering, University of Calgary, Calgary, AB, Canada. 6. McCaig Institute for Bone and Joint Health, Calgary, AB, Canada. 7. Department of Mechanical Engineering, University of British Columbia, Vancouver, BC, Canada.
Abstract
PURPOSE: C-arms are portable X-ray devices used to generate radiographic images in orthopedic surgical procedures. Evidence suggests that scouting images, which are used to aid in C-arm positioning, result in increased operation time and excess radiation exposure. C-arms are also primarily used qualitatively to view images, with limited quantitative functionality. Various techniques have been proposed to improve positioning, reduce radiation exposure, and provide quantitative measuring tools, all of which require accurate C-arm position tracking. While external stereo camera systems can be used for this purpose, they are typically considered too obtrusive. This paper therefore presents the development and verification of a low-profile, real-time C-arm base-tracking system using computer vision techniques. METHODS: The proposed tracking system, called OPTIX (On-board Position Tracking for Intraoperative X-rays), uses a single downward-facing camera mounted to the base of a C-arm. Relative motion tracking and absolute position recovery algorithms were implemented to track motion using the visual texture in operating room floors. The accuracy of the system was evaluated in a simulated operating room mounted on a real C-arm. RESULTS: The relative tracking algorithm measured relative translation position changes with errors of less than 0.75% of the total distance travelled, and orientation with errors below 5% of the cumulative rotation. With an error-correction step incorporated, OPTIX achieved C-arm repositioning with translation errors of less than [Formula: see text] mm and rotation errors of less than [Formula: see text]. A display based on the OPTIX measurements enabled consistent C-arm repositioning within 5 mm of a previously stored reference position. CONCLUSION: The system achieved clinically relevant accuracies and could result in a reduced need for scout images when re-acquiring a previous position. We believe that, if implemented in an operating room, OPTIX has the potential to reduce both operating time and harmful radiation exposure to patients and surgical staff.
PURPOSE: C-arms are portable X-ray devices used to generate radiographic images in orthopedic surgical procedures. Evidence suggests that scouting images, which are used to aid in C-arm positioning, result in increased operation time and excess radiation exposure. C-arms are also primarily used qualitatively to view images, with limited quantitative functionality. Various techniques have been proposed to improve positioning, reduce radiation exposure, and provide quantitative measuring tools, all of which require accurate C-arm position tracking. While external stereo camera systems can be used for this purpose, they are typically considered too obtrusive. This paper therefore presents the development and verification of a low-profile, real-time C-arm base-tracking system using computer vision techniques. METHODS: The proposed tracking system, called OPTIX (On-board Position Tracking for Intraoperative X-rays), uses a single downward-facing camera mounted to the base of a C-arm. Relative motion tracking and absolute position recovery algorithms were implemented to track motion using the visual texture in operating room floors. The accuracy of the system was evaluated in a simulated operating room mounted on a real C-arm. RESULTS: The relative tracking algorithm measured relative translation position changes with errors of less than 0.75% of the total distance travelled, and orientation with errors below 5% of the cumulative rotation. With an error-correction step incorporated, OPTIX achieved C-arm repositioning with translation errors of less than [Formula: see text] mm and rotation errors of less than [Formula: see text]. A display based on the OPTIX measurements enabled consistent C-arm repositioning within 5 mm of a previously stored reference position. CONCLUSION: The system achieved clinically relevant accuracies and could result in a reduced need for scout images when re-acquiring a previous position. We believe that, if implemented in an operating room, OPTIX has the potential to reduce both operating time and harmful radiation exposure to patients and surgical staff.
Entities:
Keywords:
C-arm; Computer vision; Orthopedic; Position tracking; Radiation; Tracked C-arm
Authors: Xi Qin; Bohan Wang; David Boegner; Brandon Gaitan; Yingning Zheng; Xian Du; Yu Chen Journal: IEEE Trans Biomed Eng Date: 2022-03-18 Impact factor: 4.756