Tissue Modeling in a Patient-Specific Skull Base Surgical Simulator

Abstract

Objective: Patient-specific surgical simulation is a new frontier that promises great benefits for surgical planning and rehearsal. However, it requires an efficient means of generating virtual anatomic models. Our objective is to develop a workflow to convert clinical imaging data into computational representations that support real-time simulation and haptic interaction.

Method: Standard computed tomography was obtained from 3 otologic patients. Key anatomy was segmented, cleaned by morphological operations, and converted into 3D polygonal meshes. These were reduced to the desired complexity and converted to solid tetrahedral meshes. Results were imported into the Simulation Open Framework Architecture (SOFA) for interactive simulation.

Results: Using the workflow developed, we were able to create accurate patient-specific virtual models of surgically relevant otologic anatomy. The resulting structures can be deformed and manipulated interactively through a haptic interface. Although still labor-intensive, the workflow promises considerable time-saving over purely manual segmentation and model creation. The resulting models are amenable to derivation and application of physical properties of specific tissues to improve behavioral realism.

Conclusion: Our prototype is an effective means of converting preoperative imaging into patient-specific 3D models suitable for interactive physical simulation. We continue to make model creation more automated. The potential for adding tissue-specific physical properties will augment realism and increase the utility of the simulation for surgical rehearsal.