<strong>Background:</strong><span style="font-family:;" "=""><span style="font-family:Verdana;"> Intraoperative surgical planning tools (ISPTs) used in curren...<strong>Background:</strong><span style="font-family:;" "=""><span style="font-family:Verdana;"> Intraoperative surgical planning tools (ISPTs) used in current-generation robotic arm-assisted total knee arthroplasty (RTKA) systems (such as Navio</span><sup><span style="font-size:12px;font-family:Verdana;"><span lang="ZH-CN" style="font-size:12pt;font-family:宋体;">®</span></span></sup><span style="font-family:Verdana;"> and MAKO</span><sup><span style="font-size:12px;font-family:Verdana;"><span lang="ZH-CN" style="font-size:12pt;font-family:宋体;">®</span></span></sup><span style="font-family:Verdana;">) involve employment of postoperative passive joint balancing. This results in improper ligament tension, which may negatively impact joint stability, which, in turn, may adversely affect patient function after TKA. </span><b><span style="font-family:Verdana;">Methods:</span></b><span style="font-family:Verdana;"> A simulation-enhanced ISPT (SEISPT) that provides insights relating to postoperative active joint mechanics was developed. This involved four steps: 1) validation of a multi-body musculoskeletal model;2) optimization of the validated model;3) use of the validated and optimized model to derive knee performance equations (KPEs), which are equations that relate implant component characteristics to implant component biomechanical responses;and 4) optimization of the KPEs with respect to these responses. In a proof-of-concept study, KPEs that involved two</span></span><span style="font-family:Verdana;"> </span><span style="font-family:Verdana;">com</span><span style="font-family:Verdana;">- </span><span style="font-family:;" "=""><span style="font-family:Verdana;">ponent biomechanical responses that have been shown to strongly correlate with poor proprioception (a common patient complaint post-TKA) were used to calculate optimal positions and orientations of the femoral and tibial components in the TKA design implanted in one subject (as reported in a publicly-available dataset). </span><b><span style="font-family:Verdana;">Results:</span></b><span style="font-family:Verdana;"> The differences between the calculated implant positions and orientations and the corresponding achieved values for the implant components in the subject were not similar to component position and orientation errors reported in biomechanical literature studies involving Navio</span><sup><span style="font-size:12px;font-family:Verdana;"><span lang="ZH-CN" style="font-size:12pt;font-family:宋体;">®</span></span></sup><span style="font-family:Verdana;"> and MAKO</span><sup><span style="font-size:12px;font-family:Verdana;"><span lang="ZH-CN" style="font-size:12pt;font-family:宋体;">®</span></span></sup><span style="font-family:Verdana;">. Also, we indicate how SEISPT could be incorporated into the surgical workflow of Navio</span><sup><span style="font-size:12px;font-family:Verdana;"><span lang="ZH-CN" style="font-size:12pt;font-family:宋体;">®</span></span></sup><span style="font-family:Verdana;"> with minimal disruption and increase in cost. </span><b><span style="font-family:Verdana;">Conclusion:</span></b><span style="font-family:Verdana;"> SEISPT is a plausible alternative to current-gen</span></span><span style="font-family:Verdana;">- </span><span style="font-family:Verdana;">eration ISPTs.</span>展开更多
One of the most important problems in robot kinematics and control is, finding the solution of Inverse Kinematics. Inverse kinematics computation has been one of the main problems in robotics research. As the Complexi...One of the most important problems in robot kinematics and control is, finding the solution of Inverse Kinematics. Inverse kinematics computation has been one of the main problems in robotics research. As the Complexity of robot increases, obtaining the inverse kinematics is difficult and computationally expensive. Traditional methods such as geometric, iterative and algebraic are inadequate if the joint structure of the manipulator is more complex. As alternative approaches, neural networks and optimal search methods have been widely used for inverse kinematics modeling and control in robotics This paper proposes neural network architecture that consists of 6 sub-neural networks to solve the inverse kinematics problem for robotics manipulators with 2 or higher degrees of freedom. The neural networks utilized are multi-layered perceptron (MLP) with a back-propagation training algorithm. This approach will reduce the complexity of the algorithm and calculation (matrix inversion) faced when using the Inverse Geometric Models implementation (IGM) in robotics. The obtained results are presented and analyzed in order to prove the efficiency of the proposed approach.展开更多
针对非结构化环境中多目标抓取检测存在速度慢、效果差的问题,提出一种先目标检测后抓取检测的方法。在目标检测中,为了加快网络的运行速度,文中采用深度可分离卷积和坐标注意力机制对YOLOv5网络进行轻量化改进。在抓取任务中,设计了一...针对非结构化环境中多目标抓取检测存在速度慢、效果差的问题,提出一种先目标检测后抓取检测的方法。在目标检测中,为了加快网络的运行速度,文中采用深度可分离卷积和坐标注意力机制对YOLOv5网络进行轻量化改进。在抓取任务中,设计了一种单阶段抓取位姿检测算法。首先,考虑到非结构化环境中存在的干扰,选用RGB-D图像作为抓取网络的输入数据,并选用GG-CNN作为主干网络;其次,为了加强抓取网络的特征提取能力,利用Inception-ResNet模块中不同大小卷积核的并联使用来拓宽网络感受野,同时无参三维注意力机制的融入使得网络更专注于抓取信息特征,抑制背景噪声信息;最后,使用抓取质量评估来对抓取框进行修正,并输出置信度和最高的抓取框。实验结果表明:轻量化的目标检测网络参数量为2776708,每秒帧数(frames per second,FPS)为102;改进后的抓取检测网络在公共Cornell数据集上,取得了96.57%的准确率,FPS为54.17。这说明2个网络能部署在机械臂上并较好地实现在多目标场景下的抓取任务,可以应用到实际工业生产中。展开更多
文摘<strong>Background:</strong><span style="font-family:;" "=""><span style="font-family:Verdana;"> Intraoperative surgical planning tools (ISPTs) used in current-generation robotic arm-assisted total knee arthroplasty (RTKA) systems (such as Navio</span><sup><span style="font-size:12px;font-family:Verdana;"><span lang="ZH-CN" style="font-size:12pt;font-family:宋体;">®</span></span></sup><span style="font-family:Verdana;"> and MAKO</span><sup><span style="font-size:12px;font-family:Verdana;"><span lang="ZH-CN" style="font-size:12pt;font-family:宋体;">®</span></span></sup><span style="font-family:Verdana;">) involve employment of postoperative passive joint balancing. This results in improper ligament tension, which may negatively impact joint stability, which, in turn, may adversely affect patient function after TKA. </span><b><span style="font-family:Verdana;">Methods:</span></b><span style="font-family:Verdana;"> A simulation-enhanced ISPT (SEISPT) that provides insights relating to postoperative active joint mechanics was developed. This involved four steps: 1) validation of a multi-body musculoskeletal model;2) optimization of the validated model;3) use of the validated and optimized model to derive knee performance equations (KPEs), which are equations that relate implant component characteristics to implant component biomechanical responses;and 4) optimization of the KPEs with respect to these responses. In a proof-of-concept study, KPEs that involved two</span></span><span style="font-family:Verdana;"> </span><span style="font-family:Verdana;">com</span><span style="font-family:Verdana;">- </span><span style="font-family:;" "=""><span style="font-family:Verdana;">ponent biomechanical responses that have been shown to strongly correlate with poor proprioception (a common patient complaint post-TKA) were used to calculate optimal positions and orientations of the femoral and tibial components in the TKA design implanted in one subject (as reported in a publicly-available dataset). </span><b><span style="font-family:Verdana;">Results:</span></b><span style="font-family:Verdana;"> The differences between the calculated implant positions and orientations and the corresponding achieved values for the implant components in the subject were not similar to component position and orientation errors reported in biomechanical literature studies involving Navio</span><sup><span style="font-size:12px;font-family:Verdana;"><span lang="ZH-CN" style="font-size:12pt;font-family:宋体;">®</span></span></sup><span style="font-family:Verdana;"> and MAKO</span><sup><span style="font-size:12px;font-family:Verdana;"><span lang="ZH-CN" style="font-size:12pt;font-family:宋体;">®</span></span></sup><span style="font-family:Verdana;">. Also, we indicate how SEISPT could be incorporated into the surgical workflow of Navio</span><sup><span style="font-size:12px;font-family:Verdana;"><span lang="ZH-CN" style="font-size:12pt;font-family:宋体;">®</span></span></sup><span style="font-family:Verdana;"> with minimal disruption and increase in cost. </span><b><span style="font-family:Verdana;">Conclusion:</span></b><span style="font-family:Verdana;"> SEISPT is a plausible alternative to current-gen</span></span><span style="font-family:Verdana;">- </span><span style="font-family:Verdana;">eration ISPTs.</span>
文摘One of the most important problems in robot kinematics and control is, finding the solution of Inverse Kinematics. Inverse kinematics computation has been one of the main problems in robotics research. As the Complexity of robot increases, obtaining the inverse kinematics is difficult and computationally expensive. Traditional methods such as geometric, iterative and algebraic are inadequate if the joint structure of the manipulator is more complex. As alternative approaches, neural networks and optimal search methods have been widely used for inverse kinematics modeling and control in robotics This paper proposes neural network architecture that consists of 6 sub-neural networks to solve the inverse kinematics problem for robotics manipulators with 2 or higher degrees of freedom. The neural networks utilized are multi-layered perceptron (MLP) with a back-propagation training algorithm. This approach will reduce the complexity of the algorithm and calculation (matrix inversion) faced when using the Inverse Geometric Models implementation (IGM) in robotics. The obtained results are presented and analyzed in order to prove the efficiency of the proposed approach.
文摘针对非结构化环境中多目标抓取检测存在速度慢、效果差的问题,提出一种先目标检测后抓取检测的方法。在目标检测中,为了加快网络的运行速度,文中采用深度可分离卷积和坐标注意力机制对YOLOv5网络进行轻量化改进。在抓取任务中,设计了一种单阶段抓取位姿检测算法。首先,考虑到非结构化环境中存在的干扰,选用RGB-D图像作为抓取网络的输入数据,并选用GG-CNN作为主干网络;其次,为了加强抓取网络的特征提取能力,利用Inception-ResNet模块中不同大小卷积核的并联使用来拓宽网络感受野,同时无参三维注意力机制的融入使得网络更专注于抓取信息特征,抑制背景噪声信息;最后,使用抓取质量评估来对抓取框进行修正,并输出置信度和最高的抓取框。实验结果表明:轻量化的目标检测网络参数量为2776708,每秒帧数(frames per second,FPS)为102;改进后的抓取检测网络在公共Cornell数据集上,取得了96.57%的准确率,FPS为54.17。这说明2个网络能部署在机械臂上并较好地实现在多目标场景下的抓取任务,可以应用到实际工业生产中。