Research on real-time simulation technology of virtual surgery based on BP neural network

In the virtual simulation study of human legs, it is an effective method to establish a finite element model and analyze the biomechanical characteristics. However, due to the large amount of data, long solution time, and difficulty in integrating with existing systems, the finite element model is not suitable for real-time surgical training and surgical rehearsal. In order to improve the ability of the virtual surgical simulation system to carry out actual operations, the author proposes to replace the finite element model with the BP neural network model to achieve real-time biomechanical response. Combined with the existing medical robot-assisted osteopathic virtual reality simulation surgery system, a system experimental platform was constructed. Experimental results show that the BP neural network model of human legs can fully meet the real-time requirements required for surgical simulation.