Answers are grounded in this paper and cite their sources.
Demonstrates leader-follower control for a rehabilitative soft glove using cooperative sensing, indicating enhanced hand motor recovery for stroke patients.
Rehabilitative gloves with leader-follower control offer a promising approach for stroke patients with impaired hand function. However, the current integration of sensors in rehabilitative soft gloves is limited, which poses challenges in achieving position feedback and desired bending angles of actuators. Moreover, fine movement perception and recognition add complexity to leader-follower control. Here, we propose a rehabilitative soft glove system with cooperative sensing and fine gesture recognition, utilizing a carbon black (CB) strain sensor and electromyography sensor. The strain sensor has high flexibility and stretchability, and we conducted a comprehensive analysis of its electrical characteristics. It demonstrates an impressive 100% stretchability with a gauge factor of 1.14–1.34. By seamlessly integrating the strain sensors into a data glove, we successfully monitored finger movement and grip. We propose two leader-follower control methods, namely strain sensor-based position control and cooperative sensing-based fine dynamic gesture recognition control. In the former approach, we calibrated of all sensors integrated in the data glove and actuators, enabling position control of the rehabilitative soft glove through an incremental PID controller. For fine dynamic gesture recognition, we introduce a ResNet-ConvLSTM-self-attention (RCSA) deep learning framework that eliminates the need for sensor calibration and achieves an impressive recognition accuracy of 96.20%. Finally, we validate the effectiveness of our proposed control scheme through online experiments with the rehabilitative soft glove.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: