Enhanced grip force estimation in robotic surgery: A sparrow search algorithm-optimized backpropagation neural network approach

Abstract

The absence of an effective gripping force feedback mechanism in minimally invasive surgical robot systems impedes physicians’ ability to accurately perceive the force between surgical instruments and human tissues during surgery, thereby increasing surgical risks. To address the challenge of integrating force sensors on minimally invasive surgical tools in existing systems, a clamping force prediction method based on mechanical clamp blade motion parameters is proposed. The interrelation between clamping force, displacement, compression speed, and the contact area of the clamp blade indenter was analyzed through compression experiments conducted on isolated pig kidney tissue. Subsequently, a prediction model was developed using a backpropagation (BP) neural network optimized by the Sparrow Search Algorithm (SSA). This model enables real-time prediction of clamping force, facilitating more accurate estimation of forces between instruments and tissues during surgery. The results indicate that the SSA-optimized model outperforms traditional BP networks and genetic algorithm-optimized (GA) BP models in terms of both accuracy and convergence speed. This study not only provides technical support for enhancing surgical safety and efficiency, but also offers a novel research direction for the design of force feedback systems in minimally invasive surgical robots in the future.