Jing Qin, Presenter: Nothing to Disclose

Wai-Man Pang MSc, Abstract Co-Author: Nothing to Disclose

Yim Pan Chui MPH, Abstract Co-Author: Nothing to Disclose

Tien Tsin Wong, Abstract Co-Author: Nothing to Disclose

Wai Poon MD,PHD, Abstract Co-Author: Nothing to Disclose

Pheng Ann Heng PhD, Abstract Co-Author: Nothing to Disclose

To mimic fast and realistic bleeding effect in virtual surgical training systems with the state-of-the-art hardware acceleration.

To speed up bleeding simulation, we exploit the power of physics processing unit (PPU) for computing complicated fluid dynamics of blood, and deploy the geometry shader of graphics processing unit (GPU) for rendering. Smoothed particles hydrodynamics (SPH) is used to resolve the Navier-Stokes equations that govern fluid mechanics. Human blood properties obtained from biomechanics literatures are fit into the SPH computations. To mimic non-Newtonian behavior of blood, we use a dynamic viscosity adjustment in regular time steps. Realistic blood rendering is achieved by a GPU-based marching cube (MC) algorithm. The implementation is OpenGL-based.

Experiments are performed on a PC (Intel P4 Dual Core 3.2GHz, 4GB RAM, GPU: NVIDIA GFX 8800, PPU: AGEIA PhysXP1). The new geometry shader in latest GPU suits the highly parallelizable nature of MC, an interactive rendering can be achieved. For a rendering of around 100,000 triangles, GPU-based MC rendering is about 6 times faster than that of CPU-based one. When the number of particles is around 5000, frame/second of PPU-based SPH is over 50 while that of CPU-based one drops below 10. Our bleeding simulation has been integrated with a virtual reality (VR) based orthopedics open surgery simulator, bleeding effects like tricking, flowing, or polling can be simulated. 12 novices are invited to compare our bleeding simulation with that rendered with simple particle systems; users find our realistic bleeding can improve their sense of involvement with no noticeable decrease in interactivity.

Fast and realistic bleeding simulation is computational expensive, while exploiting computation advantages from specific-purpose hardware is shown to be effective in improving simulation performance.

Our work provides a real-time blood simulation software component which is essential for VR-based surgical training systems.

Qin, J, Pang, W, Chui, Y, Wong, T, Poon, W, Heng, P, Real-time Human Blood Simulation for Medical Training of Virtual Open Surgery.  Radiological Society of North America 2007 Scientific Assembly and Annual Meeting, November 25 - November 30, 2007 ,Chicago IL. http://archive.rsna.org/2007/5012318.html