Using the PhysX engine for physics-based virtual surgery with force feedback.

Area of Science:

• Medical Simulation
• Computer Graphics
• Physics Simulation

Background:

• Developing realistic surgical simulators requires complex, computationally intensive physics-based models.
• High demands for real-time graphics (30 Hz) and force feedback (1 kHz) create latency challenges for current physics engines.
• Existing multi-physics engines offer advanced capabilities but struggle with the high-frequency updates needed for interactive simulation.

Purpose of the Study:

• To propose and demonstrate a multimodal surgical simulation environment using NVIDIA's PhysX physics library.
• To address the challenge of integrating low-frequency physics updates with high-frequency haptic feedback requirements.
• To enable the development of complex, realistic surgical trainers.

Main Methods:

• Developed a multimodal simulation environment leveraging NVIDIA's PhysX physics library.
• Implemented a collision handling layer to bridge PhysX outputs and haptic rendering devices.
• Integrated a bimanual haptic interface and per-pixel shaders for graphics realism.

Main Results:

• Achieved real-time tissue response and force feedback by managing differing update frequencies.
• Demonstrated the simulator's effectiveness through a case study of the laparoscopic adjustable gastric banding (LAGB) procedure.
• Enabled simulation of major LAGB steps at suitable rates.

Conclusions:

• Combining diverse modeling strategies for behavior, collision, and graphics is feasible and beneficial for surgical simulation.
• Multimodal environments are essential for creating realistic surgical trainers with complex organ geometries and tissue properties.
• The proposed approach successfully overcomes latency barriers, facilitating advanced surgical training simulations.