Transfer of learned dynamics between different surgical robots and operative configurations

Transfer of learned dynamics between different surgical robots and operative configurations
复制标题

不同手术机器人和手术配置之间学习动态的传递

DOI:
10.1007/s11548-022-02601-7
复制
发表时间:
2022
影响因子:
3
通讯作者:
Tumerdem, Ugur
Tumerdem, Ugur
中科院分区:
工程技术3区
文献类型:
--
作者:
Yilmaz, Nural;Zhang, Jintan;Kazanzides, Peter;Tumerdem, Ugur

文献摘要

参考文献

被引文献

相似文献

目的利用da芬奇Research Kit(dVRK),提出并实验验证了分布在世界各地的不同配置和机器人之间的动力学迁移学习(Xfer)。这可以扩展最近的研究,使用神经网络来估计患者侧机械手(PSM)的动态,以提供准确的外部末端执行器的力估计,通过适应不同的机器人和仪器,并在不同的配置,与施加在仪器上的额外的力,因为它们通过trocus.MethodsThe学习模型的目标是预测机器人运动过程中的内部关节扭矩。首先,在自由空间(FS)运动期间进行穷举训练,使用几种配置来包括重力效果。第二,以适应不同的设置,有限的训练数据收集,然后通过Xfer.ResultsXfer的神经网络进行更新可以适应FS网络训练一个机器人,在一个配置,与特定的仪器,提供可比的关节扭矩估计为不同的机器人,在不同的配置,使用不同的仪器,并通过套管针插入。这种方法的鲁棒性证明了多个PSM(从dVRK社区采样),仪器,配置和套管针ports.ConclusionXfer提供了显着的改善预测误差,而不需要从头开始完整的培训,是强大的机器人,运动配置,手术器械,和患者特定的设置范围广泛。
PurposeUsing the da Vinci Research Kit (dVRK), we propose and experimentally demonstrate transfer learning (Xfer) of dynamics between different configurations and robots distributed around the world. This can extend recent research using neural networks to estimate the dynamics of the patient side manipulator (PSM) to provide accurate external end-effector force estimation, by adapting it to different robots and instruments, and in different configurations, with additional forces applied on the instruments as they pass through the trocar.MethodsThe goal of the learned models is to predict internal joint torques during robot motion. First, exhaustive training is performed during free-space (FS) motion, using several configurations to include gravity effects. Second, to adapt to different setups, a limited amount of training data is collected and then the neural network is updated through Xfer.ResultsXfer can adapt a FS network trained on one robot, in one configuration, with a particular instrument, to provide comparable joint torque estimation for a different robot, in a different configuration, using a different instrument, and inserted through a trocar. The robustness of this approach is demonstrated with multiple PSMs (sampled from the dVRK community), instruments, configurations and trocar ports.ConclusionXfer provides significant improvements in prediction errors without the need for complete training from scratch and is robust over a wide range of robots, kinematic configurations, surgical instruments, and patient-specific setups.
通过学习的术中校正进行机器人力估计
DOI: 10.1109/ismr48346.2021.9661568
发表时间: 2021
期刊: International Symposium on Medical Robotics
影响因子: --
作者:
Wu, Jie Ying;Yilmaz, Nural;Tumerdem, Ugur;Kazanzides, Peter
通讯作者: Kazanzides, Peter
DOI: --
发表时间: 2004
期刊:
影响因子: --
作者:
Erik Wernholt
通讯作者: Erik Wernholt
达芬奇研究套件上基于神经网络的逆动力学识别和外力估计
DOI: 10.1109/icra40945.2020.9197445
发表时间: 2020
期刊: IEEE International Conference on Robotics and Automation (ICRA
影响因子: --
作者:
Yilmaz, Nural;Wu, Jie Ying;Kazanzides, Peter;Tumerdem, Ugur
通讯作者: Tumerdem, Ugur
DOI: --
发表时间: 2005
期刊: Proceedings of 2005 IEEE Conference on Control Applications, 2005. CCA 2005.
影响因子: --
作者:
Andrew C. Smith;K. Hashtrudi
通讯作者: K. Hashtrudi
使用经典和机器学习技术的实验机器人逆动力学识别
DOI: --
发表时间: 2016
期刊:
影响因子: --
作者:
Vinzenz Bargsten;José de Gea Fernández;Y. Kassahun
通讯作者: Y. Kassahun