血管介入手术系统参数辨识和模型修正方法研究  

Research on parameter identification and model correction method of vascular interventional surgery system

作　　者：孙丽岩 胡陟 张俊峰 SUN Liyan;HU Zhi;ZHANG Junfeng(School of Electronic and Electrical Engineering,Shanghai University of Engineering Science,Shanghai 201620,China)

机构地区：[1]上海工程技术大学电子电气工程学院,上海201620

出　　处：《智能计算机与应用》2024年第10期136-142,共7页Intelligent Computer and Applications

基　　金：国家自然基金青年科学基金(62003207);国家重点研发计划(2019YFC0119303)。

摘　　要：在血管介入手术中,医生更多依靠力反馈进行操作和判断。当前大多数介入手术机器人缺乏力反馈功能,影响系统透明度,准确进行模型参数预测是提高血管介入手术系统透明度的关键。本文在血管介入手术过程中,以质点弹簧模型为主手端的导管动力学建模,利用力触觉渲染获取力反馈数据。但在建模过程中无法确定动力学参数,无法达到力触觉渲染的准确性,文中通过以最小二乘法为参数辨识方法获得了精确的导管动力学模型参数。采用RBF神经网络算法,利用得到的动力学辨识参数,对主端模型中导管的物理参数进行修正。研究结果表明,本文提出的方法在导管动力学参数辨识中可以提高模型的预测精度,提升血管介入手术系统透明度。本文为系统模型参数辨识设计以及力反馈真实感的提升有重要理论意义和实用价值。In vascular interventional procedures,physicians are increasingly relying on force feedback to operate and judge.Most current interventional surgical robots do not have force feedback,which affects the transparency of the system,and accurate prediction of model parameters is essential to improve the transparency of vascular interventional surgical systems.During vascular interventional procedures,a mass-spring model is used to model the dynamics of the catheter at the main end of the hand,and force-touch rendering is used to obtain force feedback data.However,the kinetic parameters could not be accurately determined during the modeling process to achieve the accuracy of the force-touch rendering.By using the least squares method as methods for identifying parameters,accurate kinematic parameters of the catheter model are obtained.The RBF neural network algorithm is proposed to use the obtained kinematic discrimination parameters to correct the catheter physical parameters in the main tip model.The results show that the proposed method can improve the model prediction accuracy and system transparency in identifying the catheter dynamics parameters.It has theoretical significance and practical value for the design of system model parameter identification and the improvement of force feedback realism.

关 键 词：血管介入 参数辨识 RBF神经网络 力反馈透明度 

分 类 号：TB114.2[理学—运筹学与控制论]