Feasibility of Using Wearable EMG Armbands combined with Unsupervised Transfer Learning for Seamless Myoelectric Control

Myoelectric control uses electromyography (EMG) signals as human-originated input to enable intuitive interfaces with machines. As such, recent rehabilitation robotics employs myoelectric control to autonomously classify user intent or operation mode using machine learning. However, performance in such applications inherently suffers from the non-stationarity of EMG signals across measurement conditions. Current laboratory-based solutions rely on careful, time-consuming control of the recordings or periodic recalibration, impeding real-world deployment. We propose that robust yet seamless myoelectric control can be achieved using a low-end, easy-to-don and doff wearable EMG sensor combined with unsupervised transfer learning. Here, we test the feasibility of one such application using a consumer-grade sensor (Myo armband, 8 EMG channels @ 200 Hz) for gesture classification across measurement conditions using an existing dataset: 5 users x 10 days x 3 sensor locations. Specifically, we first train a deep neural network using Temporal-Spatial Descriptors (TSD) with labeled source data from any particular user, day, or location. We then apply the Self-Calibrating Asynchronous Domain Adversarial Neural Network (SCADANN), which automatically adjusts the trained TSD to improve classification performance for unlabeled target data from a different user, day, or sensor location. Compared to the original TSD, SCADANN improves accuracy by 12±5.2% (avg±sd), 9.6±5.0%, and 8.6±3.3% across all possible user-to-user, day-to-day, and location-to-location cases, respectively. In one best-case scenario, accuracy improves by 26% (from 67% to 93%), whereas sometimes the gain is modest (e.g., from 76% to 78%). We also show that the performance of transfer learning can be improved by using a better model trained with good (e.g., incremental) source data. We postulate that the proposed approach is feasible and promising and can be further tailored for seamless myoelectric control of powered prosthetics or exoskeletons.

Development and Application of Artificial Intelligence Multimedia Technology Based on Big Data

Artificial intelligence is an innovative enterprise theme that combines computer science, physiology, science, and other disciplines, including expert programs, technical experiments, games, native language understanding, cognitive rehabilitation, robotics, engineering, recognition, and other fields. The breakthrough growth made the arrival of the era of big data have a fundamental impact on many traditional industries and has also promoted the transformation of teaching methods and methods in the field of education. In the traditional classroom, due to the limitations of teaching resources, teaching time, and location, students are basically passively accepted by the teacher, and the teaching form is single. This will improve in the age of big data with well-developed network message. Modern distance education uses advanced technology such as computer networks, multimedia, and artificial intelligence to build a virtual teaching environment based on the network. It breaks through the limitations of teachers, teaching materials, experimental equipment, and other resources that exist under the traditional teaching model. Using educational resources of the same scale can expand educational capabilities exponentially and at the same time overcomes the unified progress of traditional teaching methods, single teaching methods, and students’ dissatisfaction. The shortcomings of differences and personalities that are not reflected are conducive to the improvement of the quality of education. It is the original intention of education to offer students special study resources and to stimulate students’ enthusiasm for learning. To enhance the current learning environment, the article builds a personalized multimedia network teaching system based on big data and streaming media technology. By recording the teacher’s class process as a video and uploading it to the multimedia network teaching system, students can watch and download related videos online for learning and also communicate online. The system can also provide personalized and intelligent recommendations for students based on student learning habits. The system is developed based on intelligent multimedia method and has the characteristics of security and stability and has done some help for network teaching of educational institutions.

Professionals' Views and Experiences of Using Rehabilitation Robotics With Stroke Survivors: A Mixed Methods Survey

Objective: To understand the reason for low implementation of clinical and home-based rehabilitation robots and their potential.Design: Online questionnaire (November 2020 and February 2021).Subjects: A total of 100 professionals in stroke rehabilitation area were involved (Physiotherapists n = 62, Occupation therapists n = 35).Interventions: Not applicable.Main Measures: Descriptive statistics and thematic content analysis were used to analyze the responses: 1. Participants' details, 2. Professionals' views and experience of using clinical rehabilitation robots, 3. Professionals' expectation and concerns of using home-based rehabilitation robots.Results: Of 100 responses, 37 had experience of rehabilitation robots. Professionals reported that patients enjoyed using them and they increased accessibility, autonomy, and convenience especially when used at home. The main emergent themes were: “aims and objectives for rehabilitation robotics,” “requirements” (functional, software, and safety), “cost,” “patient factors” (contraindications, cautions, and concerns), and “staff issues” (concerns and benefits). The main benefits of rehabilitation robots were that they provided greater choice for therapy, increased the amount/intensity of treatment, and greater motivation to practice. Professionals perceived logistical issues (ease of use, transport, and storage), cost and limited adaptability to patients' needs to be significant barriers to tier use, whilst acknowledging they can reduce staff workload to a certain extent.Conclusion: The main reported benefit of rehabilitation robots were they increased the amount of therapy and practice after stroke. Ease of use and adaptability are the key requirements. High cost and staffing resources were the main barriers.

Soft Pneumatic Exoskeleton for Wrist and Thumb Rehabilitation

A huge population of the world is suffering from various kinds of disabilities that make basic daily activities to be challenging. The use of robotics for limb rehabilitation can assist patients to recover faster and reduce therapist to patient ratio. However, the main problems with current rehabilitation robotics are the devices are bulky, complicated, and expensive. The utilization of pneumatic artificial muscles in a rehabilitation system can reduce the design complexity, thus, making the whole system light and compact. This paper presents the development of a new 2 degree of freedom (DOF) wrist motion and thumb motion exoskeleton. A light-weight 3D printed Acrylonitrile Butadiene Styrene (ABS) material is used to fabricate the exoskeleton. The system is controlled by an Arduino Uno microcontroller board that activates the relay to open and close the solenoid valve to actuate the wrist. It allows the air to flow into and out of the pneumatic artificial muscles (PAM) based on the feedback from the sliding potentiometer. The mathematical model of the exoskeleton has been formulated using the Lagrange formula. A Proportional Integral Derivative (PID) controller has been implemented to drive the wrist extension-flexion motion in achieving the desired set-points during the exercise. The results show that the exoskeleton has successfully realized the wrist and thumb movements as desired. The wrist joint tracked the desired position with a maximum steady-state error of 10% for 101.45ᵒ the set point.

Degree of muscle-and-tendon tonus effects on kinesthetic illusion in wrist joints toward advanced rehabilitation robotics

Due to increasing demand for rehabilitation and therapy for cerebrovascular diseases, patients require advanced development of medical rehabilitation robots. In our laboratory, we focus on the formation capability of the substitute neural path caused by brain plasticity using the kinesthetic illusion (KI), which is effective for therapies using robots. In KI, people perceive an illusionary limb movement without an actual movement when a vibration stimulus is applied to a limb’s tendons. In previous research, the optimal frequency that induces the maximum KI has a correlation factor of about 0.5 with the tendon’s natural frequency when a human subject is in a state of laxity. However, we do not know whether the above finding can be applied to actual rehabilitation because muscles and tendons are sometimes in tonus during rehabilitation, a state that varies the natural frequency. In this study, we investigate the correlation between the optimal and natural frequencies of tendon by systematically changing their tension to clarify the effects on the illusion induced by the muscle and the tendon when they are in tonus. We identified a negative correlation between the optimal and natural frequencies when they are in tonus, although a positive correlation appeared when they are in laxity. This result suggests that KI’s optimal frequency should be changed based on the degree of the tendon and muscle tonus. Therefore, our present findings provide a suitable vibration frequency that induces KI due to the degree of the tendon and muscle tonus during robot therapies.

Effects of Individualized Gait Rehabilitation Robotics for Gait Training on Hemiplegic Patients: Before-After Study in The Same Person

Background : Lower-limb exoskeleton robots are being widely used in gait rehabilitation training for stroke patients. However, most of the current rehabilitation robots are guided by predestined gait trajectories, which are often different from the actual gait trajectories of specific patients. One solution is to train patients using individualized gait trajectories generated from the physical parameters of patients. Hence, we aimed to explore the effect of individual gaits on energy consumption situation during gait rehabilitation training for hemiplegic patient with low-limb exoskeleton robot.Methods : 9 unilateral-hemiplegic patients were recruited. On the first day of the experiment, the 9 patients were guided by a low-limb exoskeleton robot, walking on a flat ground for 15 minutes in general gait trajectory, which was gained by CGA (clinical gait analysis) method.On the other day, the same 9 patients wore the identical robot and walked on the same flat ground for 15 minutes in individualized gait trajectory. The main physiological parameters including heart rate (HR) and peripheral capillary oxygen saturation (SpO2) were acquired via cardiotachometer and oximeter before and after the walking training. The energy consumption situation was indicated by the variation of the value of HR and SpO2 after walking training compared to before.Results : Between-group comparison shows that the individualized gait trajectory training results in lower increase in HR levels and decrease in SpO2 levels in experimenters compared to general gait trajectory training. Difference has statistical significance(p<0.05).Conclusions : Using individualized gait guidance in rehabilitation walking training can significantly improve energy efficiency for hemiplegic patient with stroke.Trial registration: Registered on 29 July 2021 at Chinese Clinical Trial Registry (ChiCTR2100049310). https://www.chictr.org.cn/edit.aspx?pid=130960\&htm=4

Lower limb rehabilitation robotics: The current understanding and technology

BACKGROUND: With the increasing rate of ambulatory disabilities and rise in the elderly population, advance methods to deliver the rehabilitation and assistive services to patients have become important. Lower limb robotic therapeutic and assistive aids have been found to improve the rehabilitation outcome. OBJECTIVE: The article aims to present the updated understanding in the field of lower limb rehabilitation robotics and identify future research avenues. METHODS: Groups of keywords relating to assistive technology, rehabilitation robotics, and lower limb were combined and searched in EMBASE, IEEE Xplore Digital Library, Scopus, Web of Science and Google Scholar database. RESULTS: Based on the literature collected from the databases we provide an overview of the understanding of robotics in rehabilitation and state of the art devices for lower limb rehabilitation. Technological advancements in rehabilitation robotic architecture (sensing, actuation and control) and biomechanical considerations in design have been discussed. Finally, a discussion on the major advances, research directions, and challenges is presented. CONCLUSIONS: Although the use of robotics has shown a promising approach to rehabilitation and reducing the burden on caregivers, extensive and innovative research is still required in both cognitive and physical human-robot interaction to achieve treatment efficacy and efficiency.