Tang Yi, Hao Duo, Cao Chengbing, Shi Ping, Yu Hongliu, Luan Xiaowei, Fang Fanfu
Institute of Rehabilitation Engineering and Technology, University of Shanghai for Science and Technology, Shanghai, 200093, China.
Department of Rehabilitation, Changhai Hospital, Naval Medical University, Shanghai, 200433, China.
Med Eng Phys. 2023 Mar;113:103961. doi: 10.1016/j.medengphy.2023.103961. Epub 2023 Feb 18.
Exoskeletons have become an important tool to help patients with upper extremity motor dysfunction in rehabilitation training and life assistance. In the study of the upper limb exoskeleton, the human glenohumeral joint will produce accompanying movement during the movement of the shoulder joint. This phenomenon causes a positional deviation between the shoulder joint and the exoskeleton, which affects the accuracy of exoskeleton-assisted human movement and the wearing comfort. Spend.
Taking the coronal adduction and abduction of the shoulder joint as the research object, the shoulder joint angle and glenohumeral joint bony motion trajectory were fitted by bi-level X-rays, and then the Ultium Motion motion capture system was used to collect the characteristic motion of the shoulder joint surface and establish a model. A back-propagation neural network with shoulder joint motion and shoulder width as input and the coronal position of the glenohumeral joint as output, finally applied the model to the Nimbot exoskeleton upper limb rehabilitation training robot to verify the effectiveness of the algorithm.
Real-time prediction of the glenohumeral joint motion trajectory was achieved, and the human-machine coupling compliance during the wearing of the upper limb exoskeleton was improved.
外骨骼已成为帮助上肢运动功能障碍患者进行康复训练和生活辅助的重要工具。在上肢外骨骼的研究中，人体肱盂关节在肩关节运动时会产生伴随运动。这种现象导致肩关节与外骨骼之间出现位置偏差，影响外骨骼辅助人体运动的准确性和穿戴舒适性。
以肩关节的冠状面内收和外展为研究对象，通过双平面X射线拟合肩关节角度和肱盂关节骨运动轨迹，然后使用Ultium Motion运动捕捉系统收集肩关节表面的特征运动并建立模型。以肩关节运动和肩宽为输入、肱盂关节的冠状面位置为输出构建反向传播神经网络，最后将该模型应用于Nimbot外骨骼上肢康复训练机器人以验证算法的有效性。
实现了对肱盂关节运动轨迹的实时预测，提高了上肢外骨骼穿戴过程中的人机耦合顺应性。
