Welcome!
To use the personalized features of this site, please log in or register.
If you have forgotten your username or password, we can help.
|
 |
Preoperative Analysis of Optimal Imaging Orientation in Fluoroscopy for Voxel-Based 2-D/3-D Registration
| |
|
Preoperative Analysis of Optimal Imaging Orientation in Fluoroscopy for Voxel-Based 2-D/3-D Registration
Yoshikazu Nakajima6, Yuichi Tamura7, Yoshinobu Sato6, Takahito Tashiro6, Nobuhiko Sugano8, Kazuo Yonenobu9, Hideki Yoshikawa8, Takahiro Ochi7 and Shinichi Tamura6
| (6) |
Division of Interdisciplinary Image Analysis, Osaka University Graduate School of Medicine, Suita, Osaka 565-0871, Japan |
| (7) |
Division of Computer Integrated Orthopaedic Surgery, Osaka University Graduate School of Medicine, Suita, Osaka 565-0871, Japan |
| (8) |
Department of Orthopaedic Surgery, Osaka University Graduate School of Medicine, Suita, Osaka 565-0871, Japan |
| (9) |
Osaka Minami National Hospital, Kawachinagano, Osaka 586-8521, Japan |
Abstract
We have developed a system for the 3-D localization of anatomical structures without the need for surgical exposure by using
multiple-view fluoroscopy images. In this paper, we describe the system and evaluate its application to the estimation of
optimal imaging orientations in fluoroscopy. For positional measurement, a voxel-based 2-D/3-D registration technique was
employed. Since the measurement condition depends on the object shape, spatial distribution of X-ray absorption, and overlap
of organs or structures (which differs at each imaging position), determining the optimal combination of fluoroscopy orientations
is significant. We propose a system for preoperative determination of the optimal imaging orentation by using the accuracy
estimation of stereo localization from single-plane localization results. In an experiment, the computation time needed was
10 hours, which was about 14 times shorter than the time required for a full search of imaging orientation combinations.
Fulltext Preview (Small, Large)
 References secured to subscribers.
|
|
|
|
|
|