Xin Chen1, Hemal Naik1, Lejing Wang1, Nassir Navab1, Pascal Fallavollita2. 1. Chair for Computer Aided Medical Procedures, Fakultät für Informatik, Technische Universität München, Munich, Germany. 2. Chair for Computer Aided Medical Procedures, Fakultät für Informatik, Technische Universität München, Munich, Germany. fallavol@in.tum.de.
Abstract
PURPOSE: The augmented reality (AR) fluoroscope augments an X-ray image by video and provides the surgeon with a real-time in situ overlay of the anatomy. The overlay alignment is crucial for diagnostic and intra-operative guidance, so precise calibration of the AR fluoroscope is required. The first and most complex step of the calibration procedure is the determination of the X-ray source position. Currently, this is achieved using a biplane phantom with movable metallic rings on its top layer and fixed X-ray opaque markers on its bottom layer. The metallic rings must be moved to positions where at least two pairs of rings and markers are isocentric in the X-ray image. The current "trial and error" calibration process currently requires acquisition of many X-ray images, a task that is both time consuming and radiation intensive. An improved process was developed and tested for C-arm calibration. METHODS: Video guidance was used to drive the calibration procedure to minimize both X-ray exposure and the time involved. For this, a homography between X-ray and video images is estimated. This homography is valid for the plane at which the metallic rings are positioned and is employed to guide the calibration procedure. Eight users having varying calibration experience (i.e., 2 experts, 2 semi-experts, 4 novices) were asked to participate in the evaluation. RESULTS: The video-guided technique reduced the number of intra-operative X-ray calibration images by 89% and decreased the total time required by 59%. CONCLUSION: A video-based C-arm calibration method has been developed that improves the usability of the AR fluoroscope with a friendlier interface, reduced calibration time and clinically acceptable radiation doses.
PURPOSE: The augmented reality (AR) fluoroscope augments an X-ray image by video and provides the surgeon with a real-time in situ overlay of the anatomy. The overlay alignment is crucial for diagnostic and intra-operative guidance, so precise calibration of the AR fluoroscope is required. The first and most complex step of the calibration procedure is the determination of the X-ray source position. Currently, this is achieved using a biplane phantom with movable metallic rings on its top layer and fixed X-ray opaque markers on its bottom layer. The metallic rings must be moved to positions where at least two pairs of rings and markers are isocentric in the X-ray image. The current "trial and error" calibration process currently requires acquisition of many X-ray images, a task that is both time consuming and radiation intensive. An improved process was developed and tested for C-arm calibration. METHODS: Video guidance was used to drive the calibration procedure to minimize both X-ray exposure and the time involved. For this, a homography between X-ray and video images is estimated. This homography is valid for the plane at which the metallic rings are positioned and is employed to guide the calibration procedure. Eight users having varying calibration experience (i.e., 2 experts, 2 semi-experts, 4 novices) were asked to participate in the evaluation. RESULTS: The video-guided technique reduced the number of intra-operative X-ray calibration images by 89% and decreased the total time required by 59%. CONCLUSION: A video-based C-arm calibration method has been developed that improves the usability of the AR fluoroscope with a friendlier interface, reduced calibration time and clinically acceptable radiation doses.
Keywords:
C-arm fluoroscopy; Calibration; Medical augmented reality; Multi modalities; Orthopedic and trauma surgery; Video guidance; Visualization
Authors: Wolfgang Birkfellner; Michael Figl; Klaus Huber; Franz Watzinger; Felix Wanschitz; Johann Hummel; Rudolf Hanel; Wolfgang Greimel; Peter Homolka; Rolf Ewers; Helmar Bergmann Journal: IEEE Trans Med Imaging Date: 2002-08 Impact factor: 10.048
Authors: George Stetten; Aaron Cois; Wilson Chang; Damion Shelton; Robert Tamburo; John Castellucci; Olaf von Ramm Journal: Acad Radiol Date: 2005-05 Impact factor: 3.173
Authors: Gabor Fichtinger; Anton Deguet; Ken Masamune; Emese Balogh; Gregory S Fischer; Herve Mathieu; Russell H Taylor; S James Zinreich; Laura M Fayad Journal: IEEE Trans Biomed Eng Date: 2005-08 Impact factor: 4.538