Biomedical Engineering Reference
In-Depth Information
Visual assessment has often been used as a standard and has recently been
subjected to validation. A self-consistency method is considered based on the
registration circuit, in which a set of three or more images are registered in
pairs. One major effort at validation is described. It involves intrapatient,
intermodality, rigid-body registration of the head for CT, MR, and PET images
and is based on a gold standard that employs bone-implanted fiducial markers.
Most validation efforts have been concentrated primarily in rigid registra-
tion. While not all the problems in this field are solved, progress has been
substantial, and considerably more is known about rigid registration than
nonrigid registration. Improved validation for rigid systems is still of vital
importance, but the greatest challenges in assessing the success of registra-
tion systems will lie in the nonrigid regime.
Improving registration accuracy is an important goal, but without a means
of validation no registration method can be accepted as a clinical tool. It is
hoped that this discussion of accuracy assessment will lead others to work to
improve available methods and to find new methods for assessing accuracy.
Such methods will accelerate progress towards improved registration sys-
tems and will make existing methods accessible to physicians and surgeons
and, ultimately, to their patients.
Acknowledgments
The author wishes to acknowledge the support of the Engineering and
Physical Sciences Research Council of the U.K. and the Vanderbilt University
Research Council, Nashville, Tennessee for their monetary support of this
work. He also wishes to thank Professor David Hawkes, Dr. Derek Hill, and
the other members of the Computational Imaging Science Group in the Divi-
sion of Radiological Sciences and Medical Engineering at Guy's, Kings', and
St. Thomas' Schools of Medicine and Dentistry, King's College, London, for
their hospitality and stimulating discussion while this work was undertaken.
References
1.
J.B.A. Maintz and M.A. Viergever. A survey of medical image registration.
Med.
Image Anal.,
2:1-36, 1998.
2.
M.Y. Sallam and K.W. Bowyer. Registration and difference analysis of corre-
sponding mammogram images.
Med. Image Anal.,
3:103-118, 1999.
3.
B.P. Poncelet, V.J. Wedeen, R.M. Weisskoff, and M.S. Cohen. Brain parenchyma
motion: measurement with cine echoplanar MR imaging.
Radiology,
185:645-651,
1992.
