Incorporation of diffusion tensor anisotropy in brain deformation models for updating preoperative images to improve image-guidance

Image-guided neurosurgery requires a method for compensating for the loss in registration accuracy as the surgery progresses due to tissue deformation subsequent to craniotomy at the start of a case. Modeling methods may be able to update the preoperative scans by estimating the tissue movement as surgery progresses. For these techniques to be useful, knowledge of the intrinsic mechanical and hydrodynamical tissue property parameters will be important. Preoperative DTI may be able to supply patient-specific data on model property parameters influenced by anisotropy in the flow field patterns. In this paper, we show simulation results on a model constructed from an actual clinical exam where DTI data was available. Specifically, we describe a simple technique for extracting the tensor map embodied in the DTI images and applying it within the model during tissue deformation calculations. The computed pressure field shows signs of localized disturbances which are congruent with zones having a high degree of anisotropy.