Biomedical Engineering Reference
In-Depth Information
Performing Brain Image Warping
Using the Deformation Field Predicted
by a Biomechanical Model
Grand Roman Joldes, Adam Wittek, Simon K. Warfield, and Karol Miller
Abstract Biomechanical modeling has become a viable alternative to purely
image-based approaches for predicting brain deformation during surgery. Most of
the time, a finite element mesh is used for computing the deformation field.
Although many papers discuss methods for obtaining the deformation field, there
is little information on how it will be used, especially for updating images
intraoperatively. In this paper, we discuss some requirements related to the use
of this deformation field for warping high quality preoperative brain images. A
software implementation is presented, which satisfies most of these requirements.
Based on this implementation, we outline some of the difficulties in performing
brain registration intraoperatively in real time and propose possible solutions.
1
Introduction
When performing brain surgery, the surgeon must know the precise location of all
important anatomical structures in order to carry out the intervention with the
required accuracy. As only the preoperative anatomy of the patient is known
precisely from medical images (usually Magnetic Resonance Imaging—MRI), it
is recognized that the ability to predict soft organ deformation during surgery is the
main problem in performing reliable surgery on the brain [
1
].
G.R. Joldes (
*
) • A. Wittek • K. Miller
Intelligent Systems for Medicine Laboratory, School of Mechanical
and Chemical Engineering, The University of Western Australia, 35 Stirling Highway,
Crawley 6009, WA, Australia
e-mail:
grandj@mech.uwa.edu.au
S.K. Warfield
Computational Radiology Laboratory, Department of Radiology, Children's Hospital,
Harvard Medical School, 300 Longwood Avenue, Boston, MA 02115, USA
