Image-to-Mesh Conversion for Arteriovenous Malformation Surgical Simulators

In this paper we present our effort to build a push button Image-to-Mesh conversion software for blood vessels pertinent to Arteriovenous Malformations (AVM). AVMs are tangled bundles of abnormal vessels. Cerebral AVMs pose a threat of hemorrhage that could cause severe morbidity or death. Due to the risk and complexity of AVM surgery, neurosurgeons need to be highly trained. A realistic simulator will significantly improve the training process of allowing surgeons to have hands-on experience without jeopardizing patients health. However, interactive surgery simulation has made few inroads in neurosurgery due to many challenges. One of those challenges is insufficiently descriptive anatomic modeling (i.e., mesh generation) of the brain and particularly critical tissues such as blood vessels. First, we will describe the requirements analysis for AVM simulations and then the extensions of an existing general purpose mesh generation method we developed [1], for deformable registration of brain images from tumor resection using image guided neurosurgery. Our approach is based on a two-step procedure: (1) the creation of a Body-Centered Cubic (BCC) which creates a uniform structured BCC lattice and (2) the Mesh Compression step is used to smooth the BCC mesh in order to match the tissue(s) boundary. The BCC lattice is refined according to local feature size of labels in the 3D segmented image. The resolution of each tissue is automatically adjusted based on the user-defined fidelity (0, 1]. We will conclude with preliminary results and a comparison using: (i) a similar open source mesh generation method [2] and (ii) a Delaunay-based method [3], for general purpose image-to-mesh conversion. Our preliminary results indicate that we can achieve high fidelity, good gradation and quality of the elements. However, there is a trade-off between fidelity, gradation and element quality which needs to be studied in the context of Finite Element and collisions detection computations for AVM simulations --we plan to address next. References: [1] Multi-tissue mesh generation for brain images. Y. Liu, P. Foteinos, A. Chernikov and N. Chrisochoides. In 19th International Meshing Roundtable, pp 367-384, 2010. [2] Lattice Cleaving: Conforming Tetrahedral Meshes of Multi-material Domains with Bounded Quality. J. Bronson, J. Levine, R. Whitaker, In 21st International Meshing Roundtable, pp 191-207, 2012. [3] High Quality Real-Time Image-to-Mesh Conversion for Finite Element Simulations P. Foteinos and N. Chrisochoides. In Journal on Parallel and Distributed Computing, Vol. 74, No. 2, pp 2123-2140, 2014. Acknowledgments: Research reported in this publication was supported in part by the Office of The Director, National Institutes Of Health under Award Number R44OD018334 and the NSF grants: CCF-1139864 and CCF-1439079 and by the Richard T.Cheng Endowment. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH and NSF.