Electromyography-Based Control of Lower Limb Prostheses: A Systematic Review

Myoelectric-Based Estimation of Vertical Ground Reaction Force During Unconstrained Walking by a Stacked One-Dimensional Convolutional Long Short-Term Memory Model

The volitional control of powered assistive devices is commonly performed by mapping the electromyographic (EMG) activity of the lower limb to joints' angular kinematics, which are then used as the input for regulation. However, during walking, the ground reaction force (GRF) plays a central role in the modulation of the gait, providing dynamic stability and propulsion during the stance phase. Including this information within the control loop of prosthetic devices can improve the quality of the final output, providing more physiological walking dynamics that enhances the usability and patient comfort. In this work, we explored the feasibility of the estimation of the ground reaction force vertical component (VGRF) by using only the EMG activities of the thigh and shank muscles. We compared two deep learning models in three experiments that involved different muscular configurations. Overall, the outcomes show that the EMG signals could be leveraged to obtain a reliable estimation of the VGRF during walking, and the shank muscles alone represent a viable solution if a reduced recording setup is needed. On the other hand, the thigh muscles failed in providing performance enhancements, either when used alone or together with the shank muscles. The results outline the feasibility of including GRF information within an EMG-driven control scheme for prosthetic and assistive devices.

Phasor-Based Myoelectric Synergy Features: A Fast Hand-Crafted Feature Extraction Scheme for Boosting Performance in Gait Phase Recognition

Gait phase recognition systems based on surface electromyographic signals (EMGs) are crucial for developing advanced myoelectric control schemes that enhance the interaction between humans and lower limb assistive devices. However, machine learning models used in this context, such as Linear Discriminant Analysis (LDA) and Support Vector Machine (SVM), typically experience performance degradation when modeling the gait cycle with more than just stance and swing phases. This study introduces a generalized phasor-based feature extraction approach (PHASOR) that captures spatial myoelectric features to improve the performance of LDA and SVM in gait phase recognition. A publicly available dataset of 40 subjects was used to evaluate PHASOR against state-of-the-art feature sets in a five-phase gait recognition problem. Additionally, fully data-driven deep learning architectures, such as Rocket and Mini-Rocket, were included for comparison. The separability index (SI) and mean semi-principal axis (MSA) analyses showed mean SI and MSA metrics of 7.7 and 0.5, respectively, indicating the proposed approach's ability to effectively decode gait phases through EMG activity. The SVM classifier demonstrated the highest accuracy of 82% using a five-fold leave-one-trial-out testing approach, outperforming Rocket and Mini-Rocket. This study confirms that in gait phase recognition based on EMG signals, novel and efficient muscle synergy information feature extraction schemes, such as PHASOR, can compete with deep learning approaches that require greater processing time for feature extraction and classification.

Continuous neural control of a bionic limb restores biomimetic gait after amputation

For centuries scientists and technologists have sought artificial leg replacements that fully capture the versatility of their intact biological counterparts. However, biological gait requires coordinated volitional and reflexive motor control by complex afferent and efferent neural interplay, making its neuroprosthetic emulation challenging after limb amputation. Here we hypothesize that continuous neural control of a bionic limb can restore biomimetic gait after below-knee amputation when residual muscle afferents are augmented. To test this hypothesis, we present a neuroprosthetic interface consisting of surgically connected, agonist-antagonist muscles including muscle-sensing electrodes. In a cohort of seven leg amputees, the interface is shown to augment residual muscle afferents by 18% of biologically intact values. Compared with a matched amputee cohort without the afferent augmentation, the maximum neuroprosthetic walking speed is increased by 41%, enabling equivalent peak speeds to persons without leg amputation. Further, this level of afferent augmentation enables biomimetic adaptation to various walking speeds and real-world environments, including slopes, stairs and obstructed pathways. Our results suggest that even a small augmentation of residual muscle afferents restores biomimetic gait under continuous neuromodulation in individuals with leg amputation.

Interventional effect of core stability training on pain and muscle function of youth with chronic non-specific lower back pain: A randomized controlled trial

Nowadays, due to lifestyle changes, the number of young people suffering from chronic non-specific low back pain (CNLBP) is gradually increasing. The recent guidelines for the treatment of low back pain emphasize that exercise therapy is the preferred treatment method for CNLBP. This study take ordinary college male students with CNLBP as objective of the study, focused into how core stability training affected the pain and muscle function of the CNLBP of youth. Herein, 60 male subjects were randomly divided into a control group and an experimental group, and conducted a randomized control trial in the Sports Rehabilitation Laboratory of Guangxi Normal University from September to October 2023. The control group received traditional waist strength training, while the experimental group received core stability training. VAS scores, pain symptoms scores and clinical efficacy grades were evaluated. Waist muscles fitness was evaluated, including back muscle strength, the prone upper body up's static holding time, 1-min modified sit-ups' pcs, the supine abdominal curling's static holding time and the supine leg raising's static holding time. Waist movement function was also evaluated using oswestry disability index (ODI) questionnaire. Surface electromyographic (EMG) signals were collected from rectus abdominis, erector spinae and multifidus. The independent sample t-test was used to compare groups, and the paired sample t-test was used for the data comparison before and post-exercise within the group. The results of the study found that CNLBP was improved in both the experimental and control groups in the post-exercise. Compared to pre-exercise, there are significant decrease in the VAS scores (95%CI: 2.51 to 6.51, p = 0.000), pain symptoms scores (95%CI: 2.95 to 3.55, p = 0.000), waist movement function's evaluation scores for ODI (95%CI: 2.23 to 4.31, p = 0.000), rectus abdominis' IEMG values (95%CI: 2.29 to 4.39, p = 0.000), erector spinae and multifidus' IEMG values (95%CI: 2.18 to 4.45, p = 0.000) of experimental group in the post-exercise. Compared to pre-exercise, there are significant improvement in the back muscle strength (95%CI: 12.85 to 19.49, p = 0.000), the prone upper body up's static holding time (95%CI: 9.67 to 19.17, p = 0.000), the 1-min modified sit-ups' pcs (95%CI: 8.56 to 18.12, p = 0.000), the supine abdominal curling's static holding time (95%CI: 6.73 to 19.14, p = 0.000), and the supine leg raising's static holding time (95%CI: 8.21 to 18.35, p = 0.000) of experimental group in the post-exercise. In the post-exercise,there are significant lower in the VAS scores (95%CI: 1.41 to 4.98, p = 0.000), pain symptoms scores (95%CI: 1.14 to 1.79, p = 0.011), waist movement function's evaluation scores for ODI (95%CI: 1.13 to 2.25, p = 0.000), rectus abdominis' IEMG values (95%CI: 2.36 to 4.47, p = 0.000), erector spinae and multifidus' IEMG values (95%CI: 2.24 to 4.23, p = 0.017) of experimental group than those of control group. In the post-exercise, there are significant higher in the recovery rate (p = 0.000), the prone upper body up's static holding time (95%CI: 4.16 to 8.32, p = 0.008), and the supine abdominal curling's static holding time (95%CI: 3.89 to 7.44, p = 0.000) of experimental group than those of control group. Therefore, it can be concluded that core stability training is significantly effective in treating CNLBP in youth, enhancing lower back muscle function. This therapeutic effect is primarily attributed to the improvement in muscle function.

Intuitive and versatile bionic legs: a perspective on volitional control

Active lower limb prostheses show large potential to offer energetic, balance, and versatility improvements to users when compared to passive and semi-active devices. Still, their control remains a major development challenge, with many different approaches existing. This perspective aims at illustrating a future leg prosthesis control approach to improve the everyday life of prosthesis users, while providing a research road map for getting there. Reviewing research on the needs and challenges faced by prosthesis users, we argue for the development of versatile control architectures for lower limb prosthetic devices that grant the wearer full volitional control at all times. To this end, existing control approaches for active lower limb prostheses are divided based on their consideration of volitional user input. The presented methods are discussed in regard to their suitability for universal everyday control involving user volition. Novel combinations of established methods are proposed. This involves the combination of feed-forward motor control signals with simulated feedback loops in prosthesis control, as well as online optimization techniques to individualize the system parameters. To provide more context, developments related to volitional control design are touched on.

Biomechanical effects of adding an articulating toe joint to a passive foot prosthesis for incline and decline walking

Walking on sloped surfaces is challenging for many lower limb prosthesis users, in part due to the limited ankle range of motion provided by typical prosthetic ankle-foot devices. Adding a toe joint could potentially benefit users by providing an additional degree of flexibility to adapt to sloped surfaces, but this remains untested. The objective of this study was to characterize the effect of a prosthesis with an articulating toe joint on the preferences and gait biomechanics of individuals with unilateral below-knee limb loss walking on slopes. Nine active prosthesis users walked on an instrumented treadmill at a +5° incline and -5° decline while wearing an experimental foot prosthesis in two configurations: a Flexible toe joint and a Locked-out toe joint. Three participants preferred the Flexible toe joint over the Locked-out toe joint for incline and decline walking. Eight of nine participants went on to participate in a biomechanical data collection. The Flexible toe joint decreased prosthesis Push-off work by 2 Joules during both incline (p = 0.008; g = -0.63) and decline (p = 0.008; g = -0.65) walking. During incline walking, prosthetic limb knee flexion at toe-off was 3° greater in the Flexible configuration compared to the Locked (p = 0.008; g = 0.42). Overall, these results indicate that adding a toe joint to a passive foot prosthesis has relatively small effects on joint kinematics and kinetics during sloped walking. This study is part of a larger body of work that also assessed the impact of a prosthetic toe joint for level and uneven terrain walking and stair ascent/descent. Collectively, toe joints do not appear to substantially or consistently alter lower limb mechanics for active unilateral below-knee prosthesis users. Our findings also demonstrate that user preference for passive prosthetic technology may be both subject-specific and task-specific. Future work could investigate the inter-individual preferences and potential benefits of a prosthetic toe joint for lower-mobility individuals.

A Minimal and Multi-Source Recording Setup for Ankle Joint Kinematics Estimation During Walking Using Only Proximal Information From Lower Limb

In this study, a minimal setup for the ankle joint kinematics estimation is proposed relying only on proximal information of the lower-limb, i.e. thigh muscles activity and joint kinematics. To this purpose, myoelectric activity of Rectus Femoris (RF), Biceps Femoris (BF), and Vastus Medialis (VM) were recorded by surface electromyography (sEMG) from six healthy subjects during unconstrained walking task. For each subject, the angular kinematics of hip and ankle joints were synchronously recorded with sEMG signal for a total of 288 gait cycles. Two feature sets were extracted from sEMG signals, i.e. time domain (TD) and wavelet (WT) and compared to have a compromise between the reliability and computational capacity, they were used for feeding three regression models, i.e. Artificial Neural Networks, Random Forest, and Least Squares - Support Vector Machine (LS-SVM). BF together with LS-SVM provided the best ankle angle estimation in both TD and WT domains (RMSE < 5.6 deg). The inclusion of Hip joint trajectory significantly enhanced the regression performances of the model (RMSE < 4.5 deg). Results showed the feasibility of estimating the ankle trajectory using only proximal and limited information from the lower limb which would maximize a potential transfemoral amputee user's comfortability while facing the challenge of having a small amount of information thus requiring robust data-driven models. These findings represent a significant step towards the development of a minimal setup useful for the control design of ankle active prosthetics and rehabilitative solutions.

Advances in prosthetics and orthotics

Over the past years, the field of prosthetics and orthotics has seen incredible innovations that used to be perceived as science fiction. This editorial aims to shed light on such exciting developments, exploring how they are addressing the challenges faced by individuals with limb impairments and musculoskeletal conditions.