For quantitative C-arm fluoroscopy, we have developed a unified mathematical framework to tackle the issues of intra-operative calibration, pose estimation, correspondence and reconstruction, without the use of optical/electromagnetic trackers or precision-made fiducial fixtures. Our method uses randomly distributed unknown points in the imaging volume, either naturally present or induced by randomly sticking beads or other simple markers in the image pace. After these points are segmented, a high dimensional non-linear optimization computes all unknown parameters for calibration, C-arm pose, correspondence and reconstruction. Preliminary phantom experiments indicate an average C-arm tracking accuracy of 0.9 degrees and a 3D reconstruction error of 0.8 mm, with an 80 region of convergence for both the AP and lateral axes. The method appears to be sufficiently accurate for many clinical applications, and appealing since it works without any external instrumentation and does not interfere with the workspace.
|Original language||English (US)|
|Title of host publication||Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention|
|Number of pages||9|
|State||Published - 2006|
ASJC Scopus subject areas