Quantifying the intraoperative brain deformation using interventional MR imaging
Thomas Hartkens, AJ Martinzy, WA Hal, D.J. Hawkes, Daniel Rueckert, Charles L. Truwit · 2000
The increasing use of image guided surgery systems for neurosurgery has lead to considerable recent interest in quantifying brain deformation during neurosurgery, Traditional image guided neurosurgery systems deter- mine the rigid body transformation between pre-operative images and an intraoperative coordinate system (eg.: defined by an optical localiser). These systems can be very accurate, espacially if bone-implanted fiducial markers are used, provided the rigid body assumption is valid. However, recent studies have shown that the brain surface can deform by 1Omm or more underneath a craniotomy even before any resection take place. Such significant tissue deformation invalidates the rigid body assumption. Since the recent work quantifying brain deformation has concentrated on surface features, the deformation of the structures below the surface of the brain remains unknown. We demonstrate that we can quantify brain deformation throughout the brain using an automatic non-rigid registration algorithm, and that the vol- ume change per voxel can be calculated and displayed using the Jacobian of the calculated 3D deformation field. We demonstrate this on a series of 8 resection cases carried out in the interventional MR suite at the University of Minnesota.