Constructing detailed solid and smooth surfaces from voxel data for neurosurgical simulation

  • Authors:
  • Mayumi Shimizu;Yasuaki Nakamura

  • Affiliations:
  • Department of Computer and Media Technologies, Hiroshima City University, Hiroshima, Japan;Department of Computer and Media Technologies, Hiroshima City University, Hiroshima, Japan

  • Venue:
  • ICCSA'05 Proceedings of the 2005 international conference on Computational Science and Its Applications - Volume Part III
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper deals with a neurosurgical simulation system with precise volume rendering and smooth tactile sensation. In the system, the Octree based hierarchical representation of volume data with continuous tri-cubic parametric functions, called volumetric implicit functions, and smooth boundary surfaces are introduced to provide detailed solid and smooth tactile sensation in an interactive environment. The volume data represented as voxel data, which are created from CT or MRI images, are divided into sub-volume until volumetric implicit functions can approximate voxel values accurately. An Octree manages the divided volume and parameters of the implicit functions in a hierarchical manner. Furthermore, smooth boundary surfaces are constructed by fitting points on a level surface of the implicit functions. In order to render more detailed solid than voxel precision when objects are zoomed up, sub-sampled voxels are generated by using the implicit functions. As for the tactile sensation, haptic device, PHANToM, is used to actualize a smooth reaction force which is calculated by the surface normal and the distance from a position of an instrument to the nearest surface. Incision with tactile sensation can be executed by making voxels underlying the instrument transparent, when a reaction force is greater than a limit. Several experiments reveal the effectiveness of the proposed methods.