Bilateral Back Extensor Exosuit for multidimensional assistance and prevention of spinal injuries.

Lumbar spine injuries resulting from heavy or repetitive lifting remain a prevalent concern in workplaces. Back-support devices have been developed to mitigate these injuries by aiding workers during lifting tasks. However, existing devices often fall short in providing multidimensional force assistance for asymmetric lifting, an essential feature for practical workplace use. In addition, validation of device safety across the entire human spine has been lacking. This paper introduces the Bilateral Back Extensor Exosuit (BBEX), a robotic back-support device designed to address both functionality and safety concerns. The design of the BBEX draws inspiration from the anatomical characteristics of the human spine and back extensor muscles. Using a multi-degree-of-freedom architecture and serially connected linear actuators, the device's components are strategically arranged to closely mimic the biomechanics of the human spine and back extensor muscles. To establish the efficacy and safety of the BBEX, a series of experiments with human participants was conducted. Eleven healthy male participants engaged in symmetric and asymmetric lifting tasks while wearing the BBEX. The results confirm the ability of the BBEX to provide effective multidimensional force assistance. Moreover, comprehensive safety validation was achieved through analyses of muscle fatigue in the upper and the lower erector spinae muscles, as well as mechanical loading on spinal joints during both lifting scenarios. By seamlessly integrating functionality inspired by human biomechanics with a focus on safety, this study offers a promising solution to address the persistent challenge of preventing lumbar spine injuries in demanding work environments.

Avoidance of specific calibration sessions in motor intention recognition for exoskeleton-supported rehabilitation through transfer learning on EEG data.

Exoskeleton-based support for patients requires the learning of individual machine-learning models to recognize movement intentions of patients based on the electroencephalogram (EEG). A major issue in EEG-based movement intention recognition is the long calibration time required to train a model. In this paper, we propose a transfer learning approach that eliminates the need for a calibration session. This approach is validated on healthy subjects in this study. We will use the proposed approach in our future rehabilitation application, where the movement intention of the affected arm of a patient can be inferred from the EEG data recorded during bilateral arm movements enabled by the exoskeleton mirroring arm movements from the unaffected to the affected arm. For the initial evaluation, we compared two trained models for predicting unilateral and bilateral movement intentions without applying a classifier transfer. For the main evaluation, we predicted unilateral movement intentions without a calibration session by transferring the classifier trained on data from bilateral movement intentions. Our results showed that the classification performance for the transfer case was comparable to that in the non-transfer case, even with only 4 or 8 EEG channels. Our results contribute to robotic rehabilitation by eliminating the need for a calibration session, since EEG data for training is recorded during the rehabilitation session, and only a small number of EEG channels are required for model training.

Enhancing Walking Performance With a Bilateral Hip Exoskeleton Assistance in Individuals With Above-Knee Amputation.

Transfemoral amputation is a debilitating condition that leads to long-term mobility restriction and secondary disorders that negatively affect the quality of life of millions of individuals worldwide. Currently available prostheses are not able to restore energetically efficient and functional gait, thus, recently, the alternative strategy to inject energy at the residual hip has been proposed to compensate for the lack of energy of the missing leg. Here, we show that a portable and powered hip exoskeleton assisting both the residual and intact limb induced a reduction of walking energy expenditure in four individuals with above-knee amputation. The reduction of the energy expenditure, quantified using the Physiological Cost Index, was in the range [-10, -17]% for all study participants compared to walking without assistance, and between [-2, -24]% in three out of four study participants compared to walking without the device. Additionally, all study participants were able to walk comfortably and confidently with the hip exoskeleton overground at both their self-selected comfortable and fast speed without any observable alterations in gait stability. The study findings confirm that injecting energy at the hip level is a promising approach for individuals with above-knee amputation. By reducing the energy expenditure of walking and facilitating gait, a hip exoskeleton may extend mobility and improve locomotor training of individuals with above-knee amputation, with several positive implications for their quality of life.

An Embedded Electromyogram Signal Acquisition Device.

In this study, we design an embedded surface EMG acquisition device to conveniently collect human surface EMG signals, pursue more intelligent human-computer interactions in exoskeleton robots, and enable exoskeleton robots to synchronize with or even respond to user actions in advance. The device has the characteristics of low cost, miniaturization, and strong compatibility, and it can acquire eight-channel surface EMG signals in real time while retaining the possibility of expanding the channel. This paper introduces the design and function of the embedded EMG acquisition device in detail, which includes the use of wired transmission to adapt to complex electromagnetic environments, light signals to indicate signal strength, and an embedded processing chip to reduce signal noise and perform filtering. The test results show that the device can effectively collect the original EMG signal, which provides a scheme for improving the level of human-computer interactions and enhancing the robustness and intelligence of exoskeleton equipment. The development of this device provides a new possibility for the intellectualization of exoskeleton systems and reductions in their cost.

Effects of a passive upper-body exoskeleton on whole-body kinematics, leg muscle activity, and discomfort during a carrying task.

OBJECTIVE: To compare whole-body kinematics, leg muscle activity, and discomfort while performing a 10-min carrying task with and without a passive upper-body exoskeleton (CarrySuitⓇ), for both males and females. BACKGROUND: Diverse commercial passive exoskeletons have appeared on the market claiming to assist lifting or carrying task. However, evidence of their impact on kinematics, muscle activity, and discomfort while performing these tasks are necessary to determine their benefits and/or limitations. METHOD: Sixteen females and fourteen males carried a 15kg load with and without a passive exoskeleton during 10-min over a round trip route, in two non-consecutive days. Whole-body kinematics and leg muscle activity were evaluated for each condition. In addition, leg discomfort ratings were quantified before and immediately after the task. RESULTS: The gastrocnemius and vastus lateralis muscle activity remained constant over the task with the exoskeleton. Without the exoskeleton a small decrease of gastrocnemius median activation was observed regardless of sex, and a small increase in static vastus lateralis activation was observed only for females. Several differences in sagittal, frontal, and transverse movements' ranges of motion were found between conditions and over the task. With the exoskeleton, ROM in the sagittal plane increased over time for the right ankle and pelvis for both sexes, and knees for males only. Thorax ROMs in the three planes were higher for females only when using the exoskeleton. Leg discomfort was lower with the exoskeleton than without. CONCLUSION: The results revealed a positive impact on range of motion, leg muscle activity, and discomfort of the tested exoskeleton.

Active Neural Network Control for a Wearable Upper Limb Rehabilitation Exoskeleton Robot Driven by Pneumatic Artificial Muscles.

Pneumatic artificial muscle (PAM) has been widely used in rehabilitation and other fields as a flexible and safe actuator. In this paper, a PAM-actuated wearable exoskeleton robot is developed for upper limb rehabilitation. However, accurate modeling and control of the PAM are difficult due to complex hysteresis. To solve this problem, this paper proposes an active neural network method for hysteresis compensation, where a neural network (NN) is utilized as the hysteresis compensator and unscented Kalman filtering is used to estimate the weights and approximation error of the NN in real time. Compared with other inversion-based methods, the NN is directly used as the hysteresis compensator without needing inversion. Additionally, the proposed method does not require pre-training of the NN since the weights can be dynamically updated. To verify the effectiveness and robustness of the proposed method, a series of experiments have been conducted on the self-built exoskeleton robot. Compared with other popular control methods, the proposed method can track the desired trajectory faster, and tracking accuracy is gradually improved through iterative learning and updating.

A randomized cross-over study protocol to evaluate long-term gait training with a pediatric robotic exoskeleton outside the clinical setting in children with movement disorders.

Individuals with neuromuscular disorders display a combination of motor control deficits and lower limb weakness contributing to knee extension deficiency characterized by exaggerated stance phase knee flexion. There is a lack of evidence for long-term improvement of knee extension deficiency with currently available clinical treatment programs. Our previous work testing a wearable robotic exoskeleton with precisely timed assistive torque applied at the knee showed immediate increases in knee extension during walking for children with cerebral palsy, which continued to improve over an acute practice period. When we applied interleaved assistance and resistance to knee extension, we observed improvements in knee extension and increased muscle activation indicating the potential for muscle strengthening when used over time. There is a need for additional, high-quality trials to assess the impact of dosage, intensity and volume of training necessary to see persistent improvement in lower limb function for these patient populations. This randomized crossover study (ClinicalTrials.gov: NCT05726591) was designed to determine whether 12 weeks of overground gait training with a robotic exoskeleton outside of the clinical setting, following an initial in clinic accommodation period, has a beneficial effect on walking ability, muscle activity and overall motor function. Participants will be randomized to either complete the exoskeleton intervention or continue their standard therapy for 12 weeks first, followed by a crossover to the other study component. The primary outcome measure is change in peak knee extension angle during walking; secondary outcome measures include gait speed, strength, and validated clinical scales of motor function and mobility. Assessments will be completed before and after the intervention and at 6 weeks post-intervention, and safety and compliance will be monitored throughout. We hypothesize that the 12-week exoskeleton intervention outside the clinical setting will show greater improvements in study outcome measures than the standard therapy.

Implementation of a unilateral hip flexion exosuit to aid paretic limb advancement during inpatient gait retraining for individuals post-stroke: a feasibility study.

BACKGROUND: During inpatient rehabilitation, physical therapists (PTs) often need to manually advance patients' limbs, adding physical burden to PTs and impacting gait retraining quality. Different electromechanical devices alleviate this burden by assisting a patient's limb advancement and supporting their body weight. However, they are less ideal for neuromuscular engagement when patients no longer need body weight support but continue to require assistance with limb advancement as they recover. The objective of this study was to determine the feasibility of using a hip flexion exosuit to aid paretic limb advancement during inpatient rehabilitation post-stroke. METHODS: Fourteen individuals post-stroke received three to seven 1-hour walking sessions with the exosuit over one to two weeks in addition to standard care of inpatient rehabilitation. The exosuit assistance was either triggered by PTs or based on gait events detected by body-worn sensors. We evaluated clinical (distance, speed) and spatiotemporal (cadence, stride length, swing time symmetry) gait measures with and without exosuit assistance during 2-minute and 10-meter walk tests. Sessions were grouped by the assistance required from the PTs (limb advancement and balance support, balance support only, or none) without exosuit assistance. RESULTS: PTs successfully operated the exosuit in 97% of sessions, of which 70% assistance timing was PT-triggered to accommodate atypical gait. Exosuit assistance eliminated the need for manual limb advancement from PTs. In sessions with participants requiring limb advancement and balance support, the average distance and cadence during 2-minute walk test increased with exosuit assistance by 2.2 ± 3.1 m and 3.4 ± 1.9 steps/min, respectively (p < 0.017). In sessions with participants requiring balance support only, the average speed during 10-meter walk test increased with exosuit by 0.07 ± 0.12 m/s (p = 0.042). Clinical and spatiotemporal measures of independent ambulators were similar with and without exosuit (p > 0.339). CONCLUSIONS: We incorporated a unilateral hip flexion exosuit into inpatient stroke rehabilitation in individuals with varying levels of impairments. The exosuit assistance removed the burden of manual limb advancement from the PTs and resulted in improved gait measures in some conditions. Future work will understand how to optimize controller and assistance profiles for this population.

Novice Users' Brain Activity and Biomechanics Change After Practicing Walking With an Ankle Exoskeleton.

Walking with an exoskeleton represents a sophisticated interplay between human physiology and mechanical augmentation, yet understanding of cortical responses in this context remains limited. To address this gap, this study aimed to explore cortical responses during walking with an ankle exoskeleton, examining how these responses evolve with familiarity to the augmentation. Healthy participants without prior exoskeleton experience underwent EEG, EMG, and motion capture analysis while walking with exoskeleton assistance at 1.2m/s. Initially, exoskeleton-assisted walking induced significant biomechanical changes accompanied by corresponding cortical alterations, leading to increased cortical involvement. In addition, after a brief familiarization period, significant increases in alpha band cortical power were observed, indicating decreased cortical engagement. These findings hold significance for elucidating the cortical mechanisms involved in exoskeleton-assisted walking and may contribute to the development of more seamlessly integrated augmentation devices.

Design, Development, and Functional Validation of a 3D-Printed Passive Upper Limb Exoskeleton.

Motor disability in children is evident in diagnoses such as cerebral palsy, muscular dystrophy, multiple sclerosis, or spinal muscular atrophy, among others, due to altered movement and postural patterns. This becomes more evident as the child grows and can be treated with physical therapy. The effectiveness of early interventions in facilitating an improvement in daily life activities varies depending on the child's condition. In this context, the use of exoskeletons has emerged in recent years as a valuable resource for conducting more efficient therapy processes. This work describes the design (both structural and functional) and preliminary usability and functional validation of a 3D-printed passive upper limb exoskeleton. The goal is to provide clinicians with an efficient, low-cost device that is both easy to manufacture and assemble and, in a gamified environment, serves as an assistive device to physical therapy. The device features 5 degrees of freedom, enabling both a pro-gravity and an anti-gravity mode, controlled by a series of elastic bands. This gives rise to a dual operating mode, offering assistance or resistance to different arm, forearm, and shoulder-dependent movements. Usability validation conducted by exoskeleton users showed average results in all aspects rated above 3.8 out of 5, which implies levels of satisfaction between "quite satisfied" and "very satisfied". The analysis of metrics recorded during therapy, such as the Hand Path Ratio and Success Rate (capturing user movements using an inertial sensor in the gamified environment), as well as the range of motion, reveals quantifiable improvements which can be attributed to the use of the exoskeleton: the Hand Path Ratio tended to approach 1 throughout sessions in almost all the users, the Success Rate remained stable (as users consistently were capable of completing the assigned tasks), and the range of motion showed that all patients achieved improvements of more than 10 degrees in some of the tested movements). These functional validation processes involved the participation of 7 children with varying levels of upper limb neuro-motor impairments.

Overground Gait Training With a Wearable Robot in Children With Cerebral Palsy: A Randomized Clinical Trial.

Importance: Cerebral palsy (CP) is the most common developmental motor disorder in children. Robot-assisted gait training (RAGT) using a wearable robot can provide intensive overground walking experience. Objective: To investigate the effectiveness of overground RAGT in children with CP using an untethered, torque-assisted, wearable exoskeletal robot. Design, Setting, and Participants: This multicenter, single-blind randomized clinical trial was conducted from September 1, 2021, to March 31, 2023, at 5 rehabilitation institutions in Korea. Ninety children with CP in Gross Motor Function Classification System levels II to IV were randomized. Intervention: The RAGT group underwent 18 sessions of RAGT during 6 weeks, whereas the control group received standard physical therapy for the same number of sessions during the same period. Main Outcome and Measures: The primary outcome measure was the Gross Motor Function Measure 88 (GMFM-88) score. Secondary outcome measures were the GMFM-66, Pediatric Balance Scale, selective control assessment of the lower extremity, Pediatric Evaluation of Disability Inventory-Computer Adaptive Test (PEDI-CAT), 6-minute walking test scores (distance and oxygen consumption), muscle and fat mass via bioelectrical impedance analysis, and gait parameters measured via 3-dimensional analysis. All assessments were performed for all patients at baseline, at the end of the 6-week intervention, and after the 4-week follow-up. Results: Of the 90 children (mean [SD] age, 9.51 [2.48] years; 49 [54.4%] male and 41 [45.6%] female) in the study, 78 (86.7%) completed the intervention, with 37 participants (mean [SD] age, 9.57 [2.38] years; 19 [51.4%] male) and 41 participants (mean [SD] age, 9.32 [2.37] years; 26 [63.4%] male) randomly assigned to the RAGT and control groups, respectively. Changes in the RAGT group significantly exceeded changes in the control group in GMFM-88 total (mean difference, 2.64; 95% CI, 0.50-4.78), GMFM-E (mean difference, 2.70; 95% CI, 0.08-5.33), GMFM-66 (mean difference, 1.31; 95% CI, 0.01-2.60), and PEDI-CAT responsibility domain scores (mean difference, 2.52; 95% CI, 0.42-4.63), indicating independence in daily living at postintervention assessment. At the 4-week follow-up, the RAGT group showed significantly greater improvements in balance control (mean difference, 1.48; 95% CI, 0.03-2.94) and Gait Deviation Index (mean difference, 6.48; 95% CI, 2.77-10.19) compared with the control group. Conclusions and Relevance: In this randomized clinical trial, overground RAGT using a wearable robot significantly improved gross motor function and gait pattern. This new torque-assisted wearable exoskeletal robot, based on assist-as-needed control, may complement standard rehabilitation by providing adequate assistance and therapeutic support to children with CP. Trial Registration: CRIS Identifier: KCT0006273.

Positional Analysis of Assisting Muscles for Handling-Assisted Exoskeletons.

In order to better design handling-assisted exoskeletons, it is necessary to analyze the biomechanics of human hand movements. In this study, Anybody Modeling System (AMS) simulation was used to analyze the movement state of muscles during human handling. Combined with surface electromyography (sEMG) experiments, specific analysis and verification were carried out to obtain the position of muscles that the human body needs to assist during handling. In this study, the simulation and experiment were carried out for the manual handling process. A treatment group and an experimental group were set up. This study found that the vastus medialis muscle, vastus lateralis muscle, latissimus dorsi muscle, trapezius muscle, deltoid muscle and triceps brachii muscle require more energy in the process of handling, and it is reasonable and effective to combine sEMG signals with the simulation of the musculoskeletal model to analyze the muscle condition of human movement.

Safety & efficacy of a robotic hip exoskeleton on outpatient stroke rehabilitation.

OBJECTIVE: The objective of this study was to analyze the safety and efficacy of using a robotic hip exoskeleton designed by Samsung Electronics Co., Ltd., Korea, called the Gait Enhancing and Motivating System-Hip (GEMS-H), in assistance mode only with the poststroke population in an outpatient-rehabilitation setting. METHODS: Forty-one participants with an average age of 60 and average stroke latency of 6.5 years completed this prospective, single arm, interventional, longitudinal study during the COVID-19 pandemic. Significant modifications to the traditional outpatient clinical environment were made to adhere to organizational physical distancing policies as well as guidelines from the Centers for Disease Control. All participants received gait training with the GEMS-H in assistance mode for 18 training sessions over the course of 6-8 weeks. Performance-based and self-reported clinical outcomes were assessed at four time points: baseline, midpoint (after 9 training sessions), post (after 18 training sessions), and 1-month follow up. Daily step count was also collected throughout the duration of the study using an ankle-worn actigraphy device. Additionally, corticomotor excitability was measured at baseline and post for 4 bilateral lower limb muscles using transcranial magnetic stimulation. RESULTS: By the end of the training program, the primary outcome, walking speed, improved by 0.13 m/s (p < 0.001). Secondary outcomes of walking endurance, balance, and functional gait also improved as measured by the 6-Minute Walk Test (47 m, p < 0.001), Berg Balance Scale (2.93 points, p < 0.001), and Functional Gait Assessment (1.80 points, p < 0.001). Daily step count significantly improved with and average increase of 1,750 steps per day (p < 0.001). There was a 35% increase in detectable lower limb motor evoked potentials and a significant decrease in the active motor threshold in the medial gastrocnemius (-5.7, p < 0.05) after training with the device. CONCLUSIONS: Gait training with the GEMS-H exoskeleton showed significant improvements in walking speed, walking endurance, and balance in persons with chronic stroke. Day-to-day activity also improved as evidenced by increased daily step count. Additionally, corticomotor excitability changes suggest that training with this device may help correct interhemispheric imbalance typically seen after stroke. TRIAL REGISTRATION: This study is registered with ClinicalTrials.gov (NCT04285060).

Human-in-the-Loop Optimization of Knee Exoskeleton Assistance for Minimizing User's Metabolic and Muscular Effort.

Lower limb exoskeletons have the potential to mitigate work-related musculoskeletal disorders; however, they often lack user-oriented control strategies. Human-in-the-loop (HITL) controls adapt an exoskeleton's assistance in real time, to optimize the user-exoskeleton interaction. This study presents a HITL control for a knee exoskeleton using a CMA-ES algorithm to minimize the users' physical effort, a parameter innovatively evaluated using the interaction torque with the exoskeleton (a muscular effort indicator) and metabolic cost. This work innovates by estimating the user's metabolic cost within the HITL control through a machine-learning model. The regression model estimated the metabolic cost, in real time, with a root mean squared error of 0.66 W/kg and mean absolute percentage error of 26% (n = 5), making faster (10 s) and less noisy estimations than a respirometer (K5, Cosmed). The HITL reduced the user's metabolic cost by 7.3% and 5.9% compared to the zero-torque and no-device conditions, respectively, and reduced the interaction torque by 32.3% compared to a zero-torque control (n = 1). The developed HITL control surpassed a non-exoskeleton and zero-torque condition regarding the user's physical effort, even for a task such as slow walking. Furthermore, the user-specific control had a lower metabolic cost than the non-user-specific assistance. This proof-of-concept demonstrated the potential of HITL controls in assisted walking.

Evaluation of a passive arm-support exoskeleton for surgical team members: Results from live surgeries.

BACKGROUND: Musculoskeletal symptoms and injuries adversely impact the health of surgical team members and their performance in the operating room (OR). Though ergonomic risks in surgery are well-recognized, mitigating these risks is especially difficult. In this study, we aimed to assess the impacts of an exoskeleton when used by OR team members during live surgeries. METHODS: A commercial passive arm-support exoskeleton was used. One surgical nurse, one attending surgeon, and five surgical trainees participated. Twenty-seven surgeries were completed, 12 with and 15 without the exoskeleton. Upper-body postures and muscle activation levels were measured during the surgeries using inertial measurement units and electromyography sensors, respectively. Postures, muscle activation levels, and self-report metrics were compared between the baseline and exoskeleton conditions using non-parametric tests. RESULTS: Using the exoskeleton significantly decreased the percentage of time in demanding postures (>45° shoulder elevation) for the right shoulder by 7% and decreased peak muscle activation of the left trapezius, right deltoid, and right lumbar erector spinae muscles, by 7%, 8%, and 12%, respectively. No differences were found in perceived effort, and overall scores on usability ranged from "OK" to "excellent." CONCLUSIONS: Arm-support exoskeletons have the potential to assist OR team members in reducing musculoskeletal pain and fatigue indicators. To further increase usability in the OR, however, better methods are needed to identify the surgical tasks for which an exoskeleton is effective.

Exoskeleton rehabilitation robot training for balance and lower limb function in sub-acute stroke patients: a pilot, randomized controlled trial.

PURPOSE: This pilot study aimed to investigate the effects of REX exoskeleton rehabilitation robot training on the balance and lower limb function in patients with sub-acute stroke. METHODS: This was a pilot, single-blind, randomized controlled trial. Twenty-four patients with sub-acute stroke (with the course of disease ranging from 3 weeks to 3 months) were randomized into two groups, including a robot group and a control group. Patients in control group received upright bed rehabilitation (n = 12) and those in robot group received exoskeleton rehabilitation robot training (n = 12). The frequency of training in both groups was once a day (60 min each) for 5 days a week for a total of 4 weeks. Besides, the two groups were evaluated before, 2 weeks after and 4 weeks after the intervention, respectively. The primary assessment index was the Berg Balance Scale (BBS), whereas the secondary assessment indexes included the Fugl-Meyer Lower Extremity Motor Function Scale (FMA-LE), the Posture Assessment Scale for Stroke Patients (PASS), the Activities of Daily Living Scale (Modified Barthel Index, MBI), the Tecnobody Balance Tester, and lower extremity muscle surface electromyography (sEMG). RESULTS: The robot group showed significant improvements (P < 0.05) in the primary efficacy index BBS, as well as the secondary efficacy indexes PASS, FMA-LE, MBI, Tecnobody Balance Tester, and sEMG of the lower limb muscles. Besides, there were a significant differences in BBS, PASS, static eye-opening area or dynamic stability limit evaluation indexes between the robotic and control groups (P < 0.05). CONCLUSIONS: This is the first study to investigate the effectiveness of the REX exoskeleton rehabilitation robot in the rehabilitation of patients with stroke. According to our results, the REX exoskeleton rehabilitation robot demonstrated superior potential efficacy in promoting the early recovery of balance and motor functions in patients with sub-acute stroke. Future large-scale randomized controlled studies and follow-up assessments are needed to validate the current findings. CLINICAL TRIALS REGISTRATION: URL: https://www.chictr.org.cn/index.html.Unique identifier: ChiCTR2300068398.

Experiment-free exoskeleton assistance via learning in simulation.

Exoskeletons have enormous potential to improve human locomotive performance1-3. However, their development and broad dissemination are limited by the requirement for lengthy human tests and handcrafted control laws2. Here we show an experiment-free method to learn a versatile control policy in simulation. Our learning-in-simulation framework leverages dynamics-aware musculoskeletal and exoskeleton models and data-driven reinforcement learning to bridge the gap between simulation and reality without human experiments. The learned controller is deployed on a custom hip exoskeleton that automatically generates assistance across different activities with reduced metabolic rates by 24.3%, 13.1% and 15.4% for walking, running and stair climbing, respectively. Our framework may offer a generalizable and scalable strategy for the rapid development and widespread adoption of a variety of assistive robots for both able-bodied and mobility-impaired individuals.

Feature Decoupling for Multimodal Locomotion and Estimation of Knee and Ankle Angles Implemented by Multi-Model Fusion.

Many challenges exist in the study of using orthotics, exoskeletons or exosuits as tools for rehabilitation and assistance of healthy people in daily activities due to the requirements of portability and safe interaction with the user and the environment. One approach to dealing with these challenges is to design a control system that can be deployed in a portable device to identify the relationships that exist between the gait variables and gait cycle for different locomotion modes. In order to estimate the knee and ankle angles in the sagittal plane for different locomotion modes, a novel multimodal feature-decoupled kinematic estimation system consisting of a multimodal locomotion classifier and an optimal joint angle estimator is proposed in this paper. The multi-source information output from different conventional primary models are fused by assigning the non-fixed weight. To improve the performance of the primary models, a data augmentation module based on the time-frequency domain analysis method is designed. The results show that the inclusion of the data augmentation module and multi-source information fusion modules has improved the classification accuracy to 98.56% and kinematic estimation performance (PCC) to 0.904 (walking), 0.956 (running), 0.899 (stair ascent), 0.851 (stair descent), respectively. The kinematic estimation quality is generally higher for faster speed (running) or proximal joint (knee) compared to other modes and ankle. The limitations and advantages of the proposed approach are discussed. Based on our findings, the multimodal kinematic estimation system has potential in facilitating the deployment for human-in-loop control of lower-limb intelligent assistive devices.