| Abstract
|
This paper presents a new reference data s … This paper presents a new reference data set andassociated quantification methodology to assessthe accuracy of registration ofComputerized Tomography (CT) and MagneticResonance (MR) images. We alsodescribe a new semi-automatic surface-based systemfor registering and visualizing CT and MR images.We determined the registration error of our systemusing a reference data set that was obtainedfrom a cadaver in which rigid fiducial tubes wereinserted prior to imaging. Registration error wasmeasured as the distance between ananalytic expression for each fiducial tube in oneimage set and transformed samples of thecorresponding tube obtained from the other. Registration was accomplished by first identifyingsurfaces of similar anatomic structures in eachimage set. A transformation that best registeredthese structures was determined using anon-linear optimization procedure. Even thoughthe root-mean-square (RMS) distance at the registered surfaces wassimilar to that reported by other groups, we foundthat RMS distances for the tubes was significantlylarger than the final RMS distances between theregistered surfaces. We also found thatminimizing RMS distance at the skin surface didnot minimize RMS distance for the tubes. idnot minimize RMS distance for the tubes.
|
| Author
|
Paul F. Hemler +,
S +,
Y Napel +,
Thilaka S. Sumanaweera +,
Ramani Pichumani +,
Petra A. van den Elsen +,
David L. Martin +,
John Drace +,
Inder Perkash +,
John R. Adler +
|
| Bibtype
|
techreport +
|
| Institution
|
Knowledge Systems, AI Laboratory +
|
| Key
|
KSL-94-66 +
|
| Modification dateThis property is a special property in this wiki.
|
1 May 2009 14:04:25 +
|
| Month
|
October +
|
| Note
|
Medical Computer Science +
|
| Number
|
KSL-94-66 +
|
| Tag
|
Computer science +
|
| Title
|
Registration Error Quantification of a Surface-Based Multimodality Image Fusion System +
|
| Tr id
|
KSL-94-66 +
|
| Year
|
1994 +
|
| Categories |
Technical Report,
Publication,
KSL Technical Report
|
