Converging Clinical and Engineering Research on Neurorehabilitation V

Amputation of the upper limb is a debilitating condition. The clinical state of the art for restoring movement has many limitations. Our lab has explored using a wide variety of signals from the nervous system for controlling precise finger and wrist movements in real time.

The loss of a hand disrupts critical neural pathways, and severely impacts a person’s autonomy in conducting daily activities. Recent advances in bionic solutions aim to restore motor functions using electrical signals from residual nerves and muscles. Differently, here we report the first clinical implementation of a myokinetic interface, which utilizes muscle deformation sensed through passive magnetic implants. Specifically, one participant with a transradial amputation had six permanent magnets implanted in three muscles of the residual limb. A fully self-contained myokinetic prosthetic hand, housing all hardware components within the socket, was developed. The participant controlled a dexterous robotic hand in real-time using magnet localization to detect muscle deformation. In six weeks, the participant completed functional tests with scores comparable to standard of care solutions. While this initial study highlighted some conceptual and technical challenges, it demonstrates a viable new approach for intuitive human-robot interfacing.

Robotic exoskeletons for rehabilitation usually consider only references in the joint-space with cartesian parameters, like step length and foot clearance, as a consequence of the previous. This work solves the inverse kinematics problem through a meta-heuristic stochastic approach, placing step characteristics as the inputs of the problem, rather than the consequence of the joint kinematics and generating feasible joint trajectories to achieve the rehabilitation goal.

Limb amputations profoundly impact a person’s quality of life, necessitating advanced prosthetic solutions to overcome what otherwise become life-long disabilities. Despite robotic advancements, seamlessly integrating artificial limbs for intuitive control and sensory feedback remains challenging. Intuitive motor control and sensory feedback are crucial for prosthetic functionality. Techniques like Targeted Muscle Reinnervation (TMR) and Targeted Sensory Reinnervation (TSR) offer promising solutions by using reinnervated muscles and skin to mimic natural limb functions. Emerging methods such as Regenerative Peripheral Nerve Interfaces (RPNIs), Vascularized Denervated Muscle Targets (VDMTs), and Cutaneous Mechanoneural Interfaces (CMIs) face clinical implementation challenges due to their reliance on implanted electrodes. TMR and TSR enable non-invasive interfacing, but limitations in motor signal production and reinnervation maps persist. Innovative solutions for kinesthetic feedback and natural touch experiences are still under development. Ultimately, trade-offs between surgical and engineering approaches must be managed to achieve functional, user-friendly prosthetics suitable for daily life.

Recent advances in technology and surgery have significantly expanded the options for restoring functionality in individuals with upper limb amputations. Maximizing the benefits of these advancements necessitates a multidisciplinary approach, wherein rehabilitation professionals such as physiotherapists and occupational therapists play a crucial role. The aim of this chapter is to provide an overview of indications for therapy in specific interventions aimed to facilitate the optimal use of advanced prosthetic devices. General rehabilitation components encompass patient education, pain management, addressing movement restrictions, enhancing muscle strength and endurance, and correcting postural issues. Additionally, specialized interventions such as surface electromyography-based muscular activation training in selective nerve transfers and gradual loading protocols for bone-anchored prostheses are discussed to ensure successful patient outcomes.

Rehabilitation robotic devices are used to enhance the physical capacity of individuals, yet frustration and pain from repetitive tasks often lead to dropout during physical rehabilitation. Mixed Reality (MR) can be used with these devices to offer engagement and motivation. However, further investigation is needed to understand the full effects of MR in rehabilitation. This paper presents a case study analyzing the effects on gait kinematics and the presence of adverse symptoms using a Smart Walker in an immersive MR scenario. The results indicate only slight differences with MR, such as a stride length that is 0.05 m lower compared to walking without MR.

This article introduces a Mixed Reality system which integrates a virtual interface and an obstacle avoidance strategy to enhance safe navigation. The strategy utilizes a virtual Light Detection and Ranging sensor to modulate the dynamic behavior of the UFES vWalker and a visual interface with virtual elements. An experiment involving 10 participants assessed the explainability of these elements, with tasks performed both with and without the interface. Results showed that participants felt secure during tasks, though their understanding of the interface varied, while some visual cues were clear, others caused confusion.

Mixed Reality (MR) techniques are being explored to enhance human-robot interaction (HRI) in customisable setups while also increasing engagement and motivation during therapy. However, MR systems require heavy processing and cloud computing may be explored to enable its widespread use. Nevertheless, cloud-based MR faces problems such as the effect of latency upon the system. This article introduces a cloud-based system for MR-enabled HRI in overground rehabilitation and presents a pilot study to evaluate the effects of communication issues on user experience. The observed results indicate the potential of implementing such a system in clinical settings and motivate in-depth studies in the theme.

Advancements in robotics and the growing focus on assistive mobility devices have led to the development of smart walkers. These devices provide intuitive and natural assistance for human locomotion through navigation that incorporates haptic, visual, and even auditory feedback. Smart walkers are equipped with functionalities and control strategies designed to ensure user safety during locomotion. This work discusses key considerations for control and navigation strategies for smart walkers, presents results from recent experiments, and outlines future directions for these devices.

Children with developmental disorders often exhibit reduced motivation and passiveness in their playing behaviour, impacting their engagement in motor rehabilitation tasks. Robot-assisted therapy combined with virtual/augmented reality can enhance involvement and treatment outcomes. This paper discusses the initial steps in integrating a telepresence system into a rehabilitation platform for children with Cerebral Palsy. Two experiments were performed to validate the platform. The first one tested the navigation capability of the mobile robot, showing an average mapping accuracy of 0.13 m. The second experiment demonstrated synchronization between the treadmill and the mobile robot, to demonstrate the feasibility of telepresence functionalities. Future work involves further testing, system synchronization improvements, and adding image transmission from the mobile robot to the VR headset.

The objective of this work is the creation of a control algorithm for a lower limb exoskeleton that allows assistance for the performance of different lower-limb daily tasks using a state machine. Said algorithm will be customized by adapting to both the subject’s need and the environment variables. The algorithm uses impedance-controlled torque calculation fed with a dataset obtained from healthy subjects. For the testing of the algorithm, a state machine and an exoskeleton from Fourier Intelligence customized with additional sensors will be used.

Standing up is one of the most common yet most mechanically demanding activities of daily living. Mobility impairments, in particular neurological conditions, often impede individuals ability to stand up independently. Electroencephalography (EEG) from 8 bipolar electrodes over motor cortex and electromyography (EMG) from 8 bilateral lower limb muscles were acquired from four healthy participants during sit-to-stand transfers. Event- related spectral perturbation (ERSP) and corticomuscular coherence (CMC) were calculated. Alpha waves (8–12 Hz) are observed during all phases of sit-to-stand. The bipolar electrodes that provided the most information are Fz- Cz, FCz- CPz, FC6- CP6 and FCs- CPs. Limited findings suggest that the motor cortex contributes to vastus lateralis activation during sit-to-stand.

To support, and potentially improve, gait-related tasks, wearable robots (WR) are promising, especially soft-WR. To develop WR that are acceptable for their users, users need to be part of the design and development of WR. In the scope of the SWAG project, a pragmatic Human-Centered Design approach is adopted to develop a soft exosuit suitable for different applications. Accordingly, user requirements were elicited, first generally via a literature review, then specifically targeted towards SWAG via focus groups with users for each use case. Resulting user requirements feed into design and development of the exosuit, regarding the required support, modularity, donning/doffing, look and feel, and comfort. Next step is to test first mock-ups and pre-functional prototypes with users.

A decreased hand strength is a common problem in various patient populations. The Carbonhand system is a wearable soft-robotic glove that supports patients’ hand function while performing daily activities. To assess the usability and satisfaction of the next generation of Carbonhand, a usability study with five patients with persistent hand function limitations due to trauma or arthritis was conducted. Although several usability issues were identified, usability scored positive on the System Usability Scale. Glove satisfaction was positive, control unit satisfaction was slightly lower. The current findings highlight that usability testing is important for further optimization of assistive devices, with the ultimate goal of promoting technology adoption by potential end users.

T-GRIP is an innovative robotic hand exoskeleton that supports the lateral pinch grip by actuating the thumb flexion and extension. To gain insight into the user requirements to further develop the T-GRIP prototype, focus groups with potential end users, consisting of patients and healthcare professionals were conducted. Findings gave insight into problems participants encounter in daily live, assistance that is needed, preferences regarding the development of the hardware and control of T-GRIP and valuable information to develop the business case model. Human centered design empowers potential end users to have a substantial role in the product design. Focus groups are an impactful qualitative method to discover what users expect from a device.

Stroke can cause lasting neurological impairments, posing significant individual and societal burden. Wearable technology shows promise for personalised stroke rehabilitation, yet its integration remains limited. This study aimed to understand physiotherapists’ perspectives and current practices regarding wearable technology in stroke rehabilitation in the United Kingdom. This mixed-methods study utilised a Knowledge, Attitudes and Practices questionnaire, and open-ended questions to gather the perspectives of physiotherapists. Seven physiotherapists participated in this study and the themes identified from the open questions were the physiotherapist experience with wearable technology, the cost and funding, the person-centred care and further research and guidelines required . 85.7% of the participants agreed that wearables are effective in stroke rehabilitation and 71.4% are confident using wearables. Findings underscore the importance of addressing usability, cost, and safety concerns with definitive evidence and guidelines.

This work aims to develop and assess a dual-purpose ankle exoskeleton to reduce metabolic costs during walking, based on a product development process, resulting in the development of a quasi-passive exoskeleton prototype.The study started with the development of an entirely passive prototype, comprising a custom-developed exoskeleton structure designed for military purposes. A qualitative and quantitative analysis of this prototype was performed by 30 subjects belonging to the Portuguese Armed Forces. Results show that 17 of the 30 subjects were able to reduce the metabolic costs during walking, with good comfort and range of motion outcomes. However, the clutch system was unable to adapt to each user’s gait pattern.Such outcomes were integrated into a second prototype, leading to the development of a quasi-passive system. Preliminary results show that the system can actuate the ankle according to each user’s gait pattern. Comfort and range of motion also show encouraging results.

Neural diseases, such as stroke, affect the motor abilities of millions of people worldwide. Over the past few decades, numerous innovative neurorehabilitation methods have been developed to tackle this challenge. Brain-Computer Interfaces have emerged as powerful tools for neurological therapies, particularly with the advent of deep learning techniques utilizing Convolutional Neural Networks. This study proposes and evaluates an electroencephalographic signal processing method based on continuous Progressive Neural Network algorithm for decoding motor imagery during pedaling. Our approach demonstrates classification accuracy exceeding 80%, highlighting its potential for enhancing neurorehabilitation therapies, especially by mitigating the effects of catastrophic forgetting.

Brain-computer interfaces (BCIs) have been used for rehabilitation therapies with promising results in patients with stroke or Spinal Cord Injury (SCI). While well-established for upper-limb rehabilitation, BCI paradigms for the restoration of lower-limb motor function are less well-established. To establish such a paradigm, we investigated electroencephalographic (EEG) brain activation patterns during a Motor Imagery (MI)-based BCI task to operate a Functional Electrical Stimulator (FES) targeting the Tibialis Anterior (TA) muscles. The correlation of EEG spectral power modulations in the high-beta band (18–24 Hz) during MI of the right, left, and bilateral dorsiflexion movements was evaluated across all 16 EEG recording sites in two healthy participants. High correlations were observed in the central parietal region. These findings hold promise for the development of effective BCI paradigms for restoring lower-limb function.

Engineers, neuroscientists, neurosurgeons, orthopedists and physiotherapists strive to develop and optimize bidirectional neural interfaces for patients wearing artificial limbs. Despite substantial progress made in the last decades, the sensory feedback delivered to patients remains only near-natural and difficult to convey. We believe that incorporating knowledge on neural control of movement may facilitate the design of novel bionic limbs. Identifying neuromuscular adaptations after limb amputation and characterizing their motor control response toward a bionic limb may be of great benefit. Here we review and comment on the opportunity to integrate these important aspects to minimize uncertainties while developing the next generation of bionic limbs.

Artificial sensory feedback can be used to convey information about the state of a myoelectric prosthesis to its user, potentially facilitating control and improving user experience. In this study, we tested four different methods to encode wrist flexion and extension angles into vibration stimulation profiles. Fifteen able-bodied participants used a keyboard to control a virtual hand to reach the target angle, while the feedback about the angle was conveyed using only 4 vibration motors. The outcome measures were the deviation from the target position and the time efficiency. There was no significant difference between the encoding methods, but the approach where the sensation moved from the dorsal to the volar side of the forearm showed a slight trend and favorable user preference. This work is an important step towards implementing artificial proprioceptive feedback about wrist flexion and extension, which when combined with the other recently presented approaches, can lead to a feedback system that can convey a complete state of the prosthesis.

The notion that sensory feedback is fundamental in artificial limbs has permeated scientific literature in the last decade. However, sensory feedback has not been a priority for prosthetic users, and this is arguably the reason for which commercial systems do not yet provide sensory features. Research prosthetic systems with sensory feedback have shown promising results in controlled environments, but longitudinal studies in daily life are lacking. Sensory substitution has not been a popular solution and physiologically appropriate sensory feedback systems face fundamental challenges to be overcome, such as developing high resolution neural interfaces that allow for selective biomimetic stimulation of afferent fibers. Systems involving surgical reconstruction and non-invasive stimulation are likely to form the first wave of commercially available solutions, with tactile feedback being the most prominent modality. Long-term studies in daily life will be crucial to justify the costs of sensory prosthetic features.

This preliminary study introduces a novel method for providing proprioceptive feedback in prosthetic control using a compact solution with 8 vibro-motors, which is then compared to a traditional method with 12 motors splitted in 2 arrays. The novel approach employs Gaussian interpolation for distributing the stimuli across a single motor array (Compact feedback) to convey wrist rotation and hand aperture. Tested on ten able-bodied subjects using pattern recognition for controlling a 2-DOF hand, both Compact and Conventional feedback methods showed comparable accuracy and time efficiency in a target achievement task. Results indicate that the novel feedback strategy is effective and intuitive, while requiring fewer motors, thereby demonstrating potential for amputees’ application.

Phantom Limb Pain (PLP) is a severe complication following limb loss, with limited effective treatment options. In an ongoing clinical trial, 35 individuals with limb loss completed 8 weeks of at-home Phantom Motor Execution (PME) therapy. Preliminary results indicate a significant improvement for various pain-related outcome measures with 57% of participants reporting a clinically significant treatment response. Using baseline pain measures and demographic factors, we achieve model predictions of treatment responders vs. non-responders with 78.6% accuracy. These findings highlight the potential of PME therapy as an effective treatment option for PLP management and seek to determine which patients are most likely to experience a clinically significant reduction in PLP. Identifying predictive factors that contribute to a successful response to PME therapy could enable more personalized and effective treatment plans.

Significant efforts have been devoted to providing sensory feedback for upper limb amputees, with very few studies tackling the challenge of providing meaningful proprioceptive feedback. The Myokinetic Interface introduces a novel approach utilizing implanted magnets for wireless tracking of muscle tissue, thereby offering a direct assessment of their physical contraction to facilitate control of robotic prosthetic devices. Additionally, implanted magnets offer the opportunity to stimulate musculotendon proprioceptors via untethered selective vibrations. This study presents a ready-to-use robotic platform to study kinesthesia in a trans-radial amputee fitted with the myokinetic interface. Such a platform is capable of simultaneously tracking multiple moving magnets, such as those associated with muscle contractions, while subjecting them to selective and frequency-controlled vibrations through external coils. This study marks a significant advancement toward the establishment of a scientific instrument for studying kinesthesia in humans.

Microprocessor-controlled knees (MPKs) and purely active (e.g. motor-powered) prosthetic knees have several limitations that affect their versatility. MPKs lack the support needed for power-intensive activities, while active prostheses fall short in terms of quietness, naturalness, and lightness. Adopting a user-centered design approach for transfemoral amputees, we identified the requirements for an ideal knee prosthesis. The findings from this initial research were then used to develop a design concept for a semi-powered knee prosthesis named Unico. This paper discusses these two initial development phases.

There are several neurological disorders that affect the motor development of pediatric patients. These disorders side effects can be treated through robotic rehabilitation devices. Due to the high cost, price and complexity associated to exoskeletons, exosuits have spread along the neurorrehabilitation community. As a consequence, there is a growing need to look for different measurement systems that do not depend on fixed structures or fixed rotation axes. All along this extended abstract, the development of a test bench to validate the knee and ankle flexo-extension measurements taken by three inertial sensors placed on top of the exosuit is covered. Measurements were obtained directly from the comparison of reference and variable inertial sensors; and from a model based on Gaussian Processes. Knee direct measurements were accurate due to the lack of variability, but ankle direct measurements were highly inaccurate, improving considerably with Gaussian Processes.

Understanding the interaction between wearable robotics and muscular recruitment is crucial for efficient workload sharing. This paper examines previous experiments involving three assistive technologies: a tendon-driven glove, a compliant exoskeleton for overhead work, and a functional electrical stimulation prototype. The findings highlight the importance of transparent control interfaces and biomechanical optimization in enhancing the usability and effectiveness of wearable assistive devices.

This paper presents a hierarchical detection algorithm for lower limb activities: sit-to-stand (STS) transfer, walking (W), stair ascent (SA), and stair descent (SD). The algorithm employs a Finite State Machine (FSM) and uses angular position and acceleration data in the sagittal plane to identify transitions between states. Expanding prior work [1], these activities are segmented into biomechanic-based states. The system integrates a Functional Electrical Stimulator (FES) to activate muscles in response to each activity, stimulating the lower limb muscles accordingly. Electromyography activity, which characterizes muscle activation patterns, is captured using the g.USBAMP biosignal amplifier. The algorithm was implemented in a neuroprosthetic device and validated with six healthy subjects, achieving satisfactory performance in terms of latency and accuracy.

The current exoskeleton and exosuit controllers lack direct control over biological tissue parameters such as muscle force and joint torques. This paper presents a predictive framework for controlling biological Musculotendon Unit (MTU) loads using exoskeletons. It introduces novel closed-form combined human-exoskeleton ordinary differential equation (ODE) models for simulating hopping and walking motions, which are utilized in the nonlinear model Predictive Controller (NMPC) of this framework. Through simulations, the framework demonstrates its capability to maintain MTU force below predefined thresholds. Future work aims to achieve real-time control of MTU load and joint torques during dynamic activities. The proposed framework, with the integration of NMPC, holds promise for enhancing the integration of exoskeleton technology in diverse applications, offering potential improvements in assistive and rehabilitative fields.

The human musculoskeletal system can be considered as a set of different instruments in an orchestra while the local reflex control represents the players, but who is the conductor in this concert? With this analogy, a conductor should send a signal that is used by different controllers to sync them to generate a harmonic movement. Based on our previous research and inspired by biomechanical studies, the ground reaction force can be used as an informative signal to support the control of different elements, e.g., muscles, joints, or locomotor subfunctions. Our recent simulations demonstrated the ability of this approach to generate human-like walking at different speeds, which is also robust against perturbations. This model’s predictive capability for human-like locomotion indicates that the same principle could synchronize human bodies with assistive devices, such as exoskeletons. Our experimental results with the BATEX exosuit support this hypothesis, showing reduced metabolic costs (9%) while increasing the preferred walking speed (14%). This suggests that the GRF, acting as the conductor, can effectively coordinate the concert of human and robotic systems, enhancing agility and efficiency.

Low back pain is prevalent among industrial workers due to heavy lifting and poor posture, causing significant back injuries globally and affecting health and productivity. Existing back-support exoskeletons (BSEs) lack closed-loop control for musculotendon unit (MTU) or joint loads. This study proposes a framework for closed-loop control of L5/S1 joint loads, featuring a simplified BSE dynamics model, a muscle excitation predictor (MEP), and a nonlinear model predictive control (NMPC) algorithm. The simplified model matches the neuromusculoskeletal model with a correlation %0.98, and the MEP has a prediction accuracy of 0.86 ± 0.06. Future work will develop the MPC algorithm to finalize the control framework.

Breakthroughs in assistive exoskeletons have occurred in the recent decade; both active and passive devices that provide partial joint moments in the lower limbs have reduced metabolic costs during walking by assisting muscle action. Musculoskeletal simulation is highly useful in describing the interaction between assistive moments, muscle-tendon mechanics, and walking energetics. In this study, we computed optimal assistive moments in ankle plantarflexion and hip flexion that produce minimal muscle activations during walking, described the muscle energetics, and estimated the potential reduction in metabolic cost. We described with analyses of muscle-tendon mechanics and motor control how reductions in muscle activation do not always result in metabolic cost savings.

Assessing the function of the hand is crucial to select the proper rehabilitation therapy as well as monitor the recovery progress. Therefore, many tests have been developed and validated for assessing the hand dexterity. The Virtual Eggs Test differs from the other tests due to its ability to assess dexterity considering not only the speed in performing the task but also the accuracy in regulating the grasping force. Here, we present the modifications to the Virtual Eggs Test involving both the components and the protocol made to improve both its performance and usability. The ongoing clinical trials and future studies will validate the Virtual Eggs Test as an innovative assessment tool to evaluate fine and gross dexterity independently to the pathology.

Introduccion: There has been an increased presence of exoskeletons in the rehabilitation of persons with multiple sclerosis (pwMS) in recent years, however their impact on satisfaction is unknown. Objective: To assess the usability and satisfaction of the Ekso GT® in pwMS. Methods: A retrospective study with pwMS from the Foundation Against Multiple Sclerosis of Madrid (Spain) were performed. Patients’ perspective of the Ekso GT® and the rehabilitation service were assessed with the Quebec User Evaluation with Assistive Technology (QUEST 2.0) and the Client Satisfaction Questionnaire (CSQ8) questionnaires. Results: Satisfaction of 40 pwMS were collected. All the items evaluated was above 3.5 out of 5, with safety and effectiveness being the highest. Significant correlations were found between number of sessions and satisfaction with the rehabilitation service (r = 0.532; p < 0.001). Conclusions: The usability of the Ekso GT® for pwMS was high, showing the usefulness of the device.

Almost 80 million people worldwide suffer from tremor – involuntary rhythmic shaking movements – in one or more body parts. Wearable soft robotic devices hold promise as a major practical solution for active tremor suppression, when other treatments are ineffective or too invasive. However complex clinical testing procedures limit the possibility to early evaluate the potential of novel soft actuators for tremor suppression. Here we introduce an approach for rapid evaluation of emerging tremor suppression technologies. This method combines three steps: 1) reproduction of clinically relevant tremor in a robotic test-bed (“mechanical patient”) that mimics a human forearm, 2) suppression of the tremor with electro-hydraulic actuators (Peano-HASEL), and 3) validation of achieved suppression via biomechanical modeling. We show that an antagonistic pair of slim (−1 mm) and lightweight (~15 gr) HASEL actuators would be fast and strong enough to suppress tremors of the mechanical patient, achieving reductions of 76–94% for clinically relevant mild to severe tremors of frequencies between 2–8 Hz.

Exoskeleton balance support can improve the ability to counteract perturbations. The process of human adaptation to this support, however, remains unclear. Here, we assessed how able-bodied individuals adapted to balance support provided by an ankle exoskeleton during walking, specifically when counteracting forward-directed pushes at the pelvis. Activation of the balance support led to immediate and clear reductions in both Center Of Mass (COM) displacement and soleus EMG activity. Further adaptations were observed across the first 35 perturbations for COM displacement and only across the first 5 perturbations for EMG activity before reaching a stable value. These findings demonstrate that adaptation to balance support is a rapid process. These results indicate that minimal training time is required for an individual to effectively utilize exoskeleton balance support.

The growing diffusion of rehabilitation robotics in daily clinical practice has led to the inclusion of the robot as a third agent in the one-to-one physiotherapist-patient interaction typical of traditional rehabilitation. The introduction of performance-based visual FeedBack (FB) improved the effectiveness of human-robot interaction in accomplishing goal-oriented tasks and helped to bridge the lack of physical contact between therapist and patient. We aim to investigate the patient’s perception of different therapist-mediated visual FBs during robot-assisted gait therapy in individuals with spinal cord injury by means of a multimodal assessment including high-density EEG and eye tracking. Results highlight how the presence of the therapist is essential to keep the patient actively involved in the rehabilitation task, a crucial point for improving clinical outcome.

The acceptance of wearable exoskeletons and their encouraging results in the clinic have prompted patients to request more of this technology. Companies and researchers are actively working to transition these exoskeletons from clinical applications to personal use. In this work, we show the evolution of an initial knee-powered orthosis concept into a hip-knee-powered lower-limb exoskeleton designed for clinical purposes, and its subsequent adaptation for personal use. Our experience offers valuable insights that can help in future research and development of this technology.

This abstract presents three mechatronic concepts with a high potential for the design of affordable force-controlled exoskeletons. These concepts allow to reduce costs and improve exoskeleton controllability at the same time. The concepts are related to (1) how to avoid oversizing actuators to obtain more lightweight, affordable, and transparent systems (2) how to enhance force controllability by mechanics (3) how to improve device transparency.

In the study of Activities of Daily Living (ADLs), limb coordination is essential. Specifically, upper extremity bimanual tasks significantly influence human capabilities. Conditions such as nerve injuries or strokes can result in unilateral paralysis. Rehabilitative assistive devices aim to mitigate these functional impairments. This paper demonstrates the importance of bimanual tasks and their significance in developing assistive devices. Six bimanual ADLs were defined, encompassing both standing and seated tasks, various categories of ADLs, and different workspace areas. While a laboratory setting inherently has limitations in replicating realistic scenarios, representative movement patterns of daily living were identified.

To ensure that a rehabilitation exercise performed with an exoskeleton is carried out safely and effectively, it is crucial to understand the upper limb kinematics during such use. Because it is impractical to assess these using motion capture and analysis tools during rehabilitation, this abstract presents two models for estimating them based on data extracted from the exoskeleton and then compares their results to measured kinematics. Compared to previous work, one of the models provided here exhibits lower deviation from measured data with only -0.21° ± 2.73° variation when examining the elbow joint alone and -2.84° ± 7.12° deviation when examining all joints. This model appears to be more suited for our needs as it takes approximately 24 times less time to compute.

Patients with stroke can experience significant gait impairments, limiting their mobility and quality of life. Robotic exoskeletons offer a promising avenue for gait rehabilitation, but simple-to-use and effective control strategies are crucial for clinical settings. This paper presents a pilot study investigating the feasibility and potential benefits of a data-driven impedance control approach for lower-limb exoskeletons in two participants with stroke. Participants underwent a training session with the exoskeleton during which kinematic and EMG data were collected. Preliminary results from these experiments suggest that the data-driven impedance controller was able to provide appropriate assistance during five distinct ambulation modes (level-ground walking, ramp ascent/descent, stair ascent/descent), increasing stride length, foot clearance, reducing gait speed variability, while maintaining the user active in the exoskeleton (similar or higher EMG activation), in comparison to natural walking.

Interaction forces exchanged between exoskeletons and humans should be carefully monitored to avoid damage to the human musculoskeletal system. This paper presents the instrumentation of a knee exoskeleton able to measure physical human-exoskeleton interaction (pHEI) forces in three axes by strategically placing force sensors at the cuffs of the device. We used an instrumented replica of a human leg to validate the force sensor readings. Results showcased the ability of the device to accurately measure forces transmitted between the exoskeleton and the human during use of the device at various motion trajectories.

Maintaining balance is a critical aspect of human gait, especially for individuals with balance deficiencies. Accurate assessment and monitoring of balance are essential to reduce the risk of falls and improve mobility outcomes. Individuals with balance impairments tend to adopt larger step widths, resulting in increased mediolateral center of mass displacement with increased step-to-step variability. We aim to establish a quantitative metric that can accurately detect balance deficiencies, potentially improving the safety and effectiveness of lower limb exoskeletons. Using the linear inverted pendulum model, we estimated the mediolateral center of mass displacement in three healthy subjects walking with narrow, normal, and wide step widths. Our analysis demonstrates the importance of mediolateral center of mass displacement for assessing balance and stability during walking showing the significant differences in this metric with varying step widths. By leveraging this metric, we aim to enhance the balance assistance with lower limb exoskeletons.

When using lower limb exoskeletons, the human body naturally adopts neuromotor learning to adapt to the newly introduced conditions and interactions. However, due to the limited comprehension of human adaptation processes and strategies, defining the methodologies and metrics for identifying when and how the adaptation occurred remains a challenge. This study aims to investigate the human adaptation behaviour during exoskeleton use from the biomechanics perspective. Rather than focusing only on the statistical inferences of the general population, discrete analyses of individual participants are carried out to capture different adaptation strategies. The preliminary results indicated potential human-exoskeleton adaptation features exhibited by the converged increase of plantar flexion angle upon exoskeleton use. On the contrary, the three participants displayed unique response patterns in metabolic cost, suggesting that generally defining adaptation with only metabolic cost information may potentially be erroneous.

Lower limb exoskeletons play a crucial role in rehabilitation and compensation of motor functions, benefiting individuals with neuromotor pathologies or muscle weakness. This paper presents the development of a variable impedance knee exoskeleton combined with a motor neuroprosthesis (MNP). The system integrates an expert system to detect sub-phases of gait, a functional electrical stimulation (FES) system, and a magnetorheological (MR) fluid damper. The objective is to provide stability and assistance during gait through real-time algorithms that control the resistance and assistance levels of the exoskeleton.

Persons with spinal cord injury (SCI) are immobilized. Unloading of the sublesional skeleton, and an inability to ambulate, results in muscle loss and bone loss. The study evaluated the effect of 100 h (5 h per week) of exoskeleton step training for non-ambulatory SCI on bone mineral density (BMD). The results have implications for length of training period and intensity of training and overall research design for future research studies.

Wearable robotic exoskeletons are frequently explored for their efficacy in rehabilitation and in assistance in daily activities in people with motor disorders, yet relatively few have convincing evidence for use. Here we describe a cable-driven ankle exoskeleton that provides assistance to the ankle in sagittal and frontal planes simultaneously, aimed for persons with dropfoot and excessive inversion after e.g. stroke. In this study, we propose a multi-objective human-in-the-loop optimization that adjusts exoskeleton control parameters to improve two independent gait quality measures, specifically foot segment kinematics and step length symmetry. We illustrate how the identified solutions represent a balance between the two objectives.

This paper describes improvements to the EduExo Pro exoskeleton, emphasizing the introduction of impedance control with a Dynamixel servo motor and a Raspberry Pi. This upgrade enhances the system’s ability to support the development of advanced control algorithms, increasing its usefulness for educational purposes. Initial findings indicate a need for friction compensation due to the high gear reduction in the new motor setup. Nevertheless, the upgrades not only boost the functionality of the EduExo Pro but also extend its educational value, allowing the development of many different control algorithms. Future work will aim to develop a gravity compensation algorithm with the included triple-axis accelerometer to further refine the system’s performance.

This paper describes the evolution of the arm therapy robot ARMin I–IV, ChARMin, and ANYexo used for rehabilitation training of neurological patients. ARMin was validated in several clinical studies and commercialized by Hocoma AG as ArmeoPower. Later, the development of the wearable, soft, cable-driven Myoshirt started with the goal of transferring the experience obtained with the ARMin series to a device that can assist people with upper limb weakness in daily life situations.

Increasing the dose of therapy administered to persons after neurological injuries may improve their functional outcome. Unsupervised robot-assisted therapy is one possible approach to support therapy along the continuum of care without overloading the healthcare system. This work summarizes the design of ReHandyBot, an active, portable device for upper limb rehabilitation targeting unsupervised use in the clinic or at home. Additionally, its preliminary usability evaluation by four stroke subjects who used the device for two weeks is described. The results showed that ReHandyBot is a suitable platform for implementing quality robot-assisted exercises, and stroke patients could use it with minimal supervision, rating its usability as excellent (mean score for the System Usability Scale > 85). This pilot study opens the door to a larger study with ReHandyBot aiming at increasing therapy dose after stroke.

A wearable soft-robotic glove used as assistive device for hand grip in daily life has shown an additional therapy effect on grip strength, although the extent of this effect varies across individuals. Besides, clinical applicability is crucial for adoption of wearable technology in practice. Investigating its clinical applicability via interviews and questionnaires revealed that use of the soft-robotic glove is well received and accepted by both patients and healthcare professionals (HCP). To better understand which factors were associated with larger changes in grip strength, correlations were calculated. No obvious relations were found with any of the subject characteristics, nor with use time. Further research is needed to better understand the factors describing which patients are more likely to benefit from use of the soft-robotic glove.

Cerebrovascular diseases like stroke, predominantly affect older adults, and “return to work” (RTW) has rarely been studied as a rehabilitation outcome. Existing data on RTW post-stroke show wide variability, with rates ranging from 11–85%. Lack of work reintegration is linked to increased cardiac issues, depression, mortality, social isolation, and poor adaptive skills. This study evaluated the impact of technology-based rehabilitation using optic (SonicHand) and wearable (Riablo™) systems, which provide auditory and visual biofeedback, on RTW outcomes for stroke patients. Although no significant differences were found in RTW rates between technological and conventional rehabilitation, the former had a higher odds ratio (OR = 9 vs. 6). Assistive devices were more commonly used among non-RTW patients. Additionally, patients who returned to their pre-stroke employment conditions reported better quality of life.

This work presents a study investigating the use of an impedance-controlled knee orthosis to assist with sit-to-stand transitions by leveraging EMG signals. A modified knee orthosis equipped with an electric actuator and an EMG bracelet enabled EMG-based impedance control. The results show altered velocity peaks with assistive support, indicating less strenuous movement. Further, EMG muscle activity was reduced significantly, suggesting lower physical effort required for the task. These findings demonstrate the potential of smart orthoses to optimize energy efficiency and enhance rehabilitation outcomes by dynamically adjusting joint impedance based on EMG activity.

We developed a myokinetic interface for controlling robotic prostheses, which monitors muscle contraction by tracking permanent magnets implanted in the residual muscles through magnetic sensors in the prosthetic socket. In the first in-human demonstration, a trans-radial amputee used six magnets implanted in three forearm muscles to control a prosthesis, achieving performance comparable to standard myoelectric controllers. However, changes in limb position and repeated prosthetic donning/doffing caused unintended relative movements between the magnets and the socket/sensors. These movements degraded magnet localization accuracy, thus significantly increasing the csontrol signal variability and challenging the performance of pattern recognition algorithms for movement classification. To address this issue, here we investigate the influence of different data processing techniques and training schemes on the classifier performance. Results demonstrate that recalibrating the system after each donning/doffing and filtering out unintended magnet displacement caused by limb movements allow to achieve classification accuracies greater than 90%.

Accurate identification of forearm muscle activity is an open question in the field of assistive and prosthetic technology. Neurorehabilitation therapies could benefit significantly from improved intention detection algorithms. Motor control models and high-density electrodes can be combined to enhance the effectiveness of such algorithms. In this paper, data from a linear, high-density electrode grid was analyzed through muscle synergy analysis, using an optimized criterion to determine the number of synergies. The results demonstrate that reducing electrode density also reduces the number of identified synergies, supporting the hypothesis of dense electrode grids are essential for comprehensive mapping of forearm muscle activity is needed.

Locomotion identification is crucial for high-level exoskeleton control, enabling real-time recognition of transitions across modes (e.g., overground walking, ramps, stairs). This paper presents an approach that relies solely on exoskeleton sensor data, combining a heuristic and a data-driven method based on a normative dataset. Preliminary results from 5 healthy participants show the method’s ability to distinguish between locomotion modes. Future work will integrate a controller that applies the desired torque to assist users in performing each movement.

Recent advancements in physical human-robot-human interaction allow people to physically interact with each other through the robotic systems that they are in contact with. This preliminary study explores a potential experimental setup using exoskeleton-mediated interactions for gait training, where a teacher provides haptic guidance to a student. The experimental setup involves haptically coupling the hip and knee joints of pilot users wearing lower-limb exoskeletons, without visual feedback on the desired gait trajectory for the student. Initial observations suggest that while the student can potentially learn a new gait pattern through physical cues alone, several challenges were noted.

Haptic human-robot-human interaction allows users to feel and respond to one another’s forces while interfacing with separate robotic devices. For both upper- and lower-limb tasks, previous work has shown that virtual interactions with a partner can improve motor performance and enhance individual learning. However, whether the mechanism of these improvements generalizes across different human systems is an open question. In this work, we investigate the effects of dyadic interaction during a trajectory tracking task involving single-joint movements at the wrist and ankle. We compare tracking performance and muscle activation during haptic conditions where pairs of participants were uni- and bi-directionally connected, in order to investigate the contribution of real-time responses from a partner during the interaction. Findings indicate similar improvements in tracking performance during the haptic conditions across joints, suggesting that uni-directional interaction is sufficient for movement correction during simple motor behaviors in healthy individuals.

As the cost of rehabilitation robots decreases, it is increasingly common for rehabilitation clinics to own multiple robots. Patients can use such robots to exercise together in a group setting, tentatively dubbed a robotic rehabilitation gym. Such gyms have potential as a new rehabilitation delivery model, as they could increase patient motivation through social interaction and reduce the number of therapists needed per patient. However, such gyms may benefit from additional software support to avoid overloading the supervising therapist and allow them to focus on the most critical aspects of therapy. This paper thus presents our ongoing work on intelligent dynamic patient-robot assignment in a simulated robotic gym via either stochastic optimization or learning from a human expert. Furthermore, we discuss possible next steps for actual real-world studies of robotic gyms.

Teleoperated robotic systems provide surgeons with precise control, reduced fatigue, and remote access to patients but have yet to break out meaningfully in gait rehabilitation despite their potential to offload the extremely taxing nature of skilled, hands-on assistance during gait training. One way to provide teleoperated rehabilitation is with coupled exoskeletons worn by patient and therapist, but this may still demand significant therapist effort and does not mimic their typical hands-on assistive strategies. Instead, end-effector robotic arms can be teleoperated with a small manipulandum at a desk to provide remote gait training assistance at an attachment point on the patient’s leg. This allows specific, targeted movements and interactions that are more congruent with therapists’ dexterity in applying forces with their hands. Here, we explore gait training with a teleoperated robotic arm by showcasing our approach and discussing the implications of augmenting therapist manipulative capabilities with end effector robotics.

The search for a physiologically appropriate control interface for prosthetic hands is an ongoing challenge in rehabilitation robotics. We proposed a myokinetic interface based on localizing magnets implanted in the residual muscles to monitor contractions and decode user intention. In the first human demonstration, we used a transcutaneous magnet localizer with a computational unit implementing magnet localization by acquiring synchronized magnetic field recordings from multiple acquisition units. These had to be arranged in fixed relative position and orientation, a constraint that substantially affected the system footprint. To overcome this limitation, here we present a modular system in which localization is implemented directly on each acquisition unit processor. Compared to the previous solution, this system achieves similar localization accuracy and precision with one or two magnets, while we devise future improvements in computation efficiency. This system significantly reduces the interface footprint and can be smoothly adapted to different anatomical districts.

This work aims to enhance shared-autonomy control methods in myoelectric hand prostheses. We present a method to predict object slipping using a machine learning algorithm, that combines friction cone theory and bandpass filtering to detect and classify slips according to the context, particularly as “safe” and “risky” slips. Preliminary experimental validation with a multi-articulated prosthesis equipped with six Hall-effect tactile sensors showed the method effectiveness in discriminating among the two slips categories. Despite this approach faced some challenges, results showed the potential to avoid autonomous reactions where there is no potential risk involved, such as social interactions (i.e. passing an object) or voluntary object releasing.

Regaining impaired sensorimotor functions is still an open challenge. Osseoperception has emerged as a potential sensory feedback modality. However, this phenomenon has been little explored upon stimulation of the distal bones in the upper limb. This study aimed to preliminarily explore osseoperception by applying vibratory stimuli at different frequencies to the pisiform bone at the wrist of four healthy participants. The lowest perception threshold (0.006 ± 0.002 N) was found at 200 Hz and the highest (0.50 ± 0.26 N) at 1500 Hz, frequencies similar to those reported in previous studies on invasive stimulation. This first insight highlights the wrist as a potential point to evoke osseoperception, which could be used to deliver augmented sensory feedback to people with neuromotor impairments.

Myoelectric control has been traditionally used in clinical applications, but from recently, there has been an increasing interest in applying this approach for more general human-machine interfacing. Here, we present NeuraLoop, a compact system for the simultaneous recording of electrical muscle activity (EMG) and delivery of electrotactile stimulation. The system uses a matrix electrode with 32 stimulation and 32 recording pads, thereby allowing high- resolution EMG recording for gesture recognition and spatially distributed stimulation for high-fidelity haptic feedback. We demonstrated the system by using NeuraLoop to detect and classify micro-gestures, which are quick, small, and transient movements, often used to interact with consumer devices. The preliminary results are encouraging although there is room for improvement. Future work will increase gesture classification performance and add haptic feedback, opening opportunities for many relevant applications of bidirectional human-machine interfacing using gesture recognition and electrotactile haptics.

Body-Machine Interfaces (BoMIs) provide a means to control various devices, enabling users to extend their motor capabilities by using the remaining redundancy in the musculoskeletal system after a neurological injury. Here, we considered a hybrid BoMI combining motion and muscle activities, measured respectively by inertial sensors and electromyography. We aimed to determine which algorithm for dimensionality reduction between a linear - principal component analysis (PCA) - and a non-linear one – nonlinear autoencoder (AE)- would allow for a more proficient control. We recruited fourteen healthy subjects and assessed their proficiency in controlling a computer cursor with either mapping. The subjects were randomly assigned to start with either PCA or AE mapping in a crossover study.We found that the hybrid BoMI with PCA led to better performance paving the way to further exploitation of linear dimensionality reduction algorithms in clinical approaches targeting simultaneously motion and muscle activations.

Technology-aided interventions and assessments are at the base of personalized motor rehabilitation after stroke.In the AVANCER trial, the effects of a novel personalized neurotech intervention were assessed by using an instrumented Fugl-Meyer (FM) evaluation pre and post interventions. The activity of 9 muscles in the affected and unaffected side of 12 severe chronic stroke patients was acquired during the FM test.Spinal maps showed specific topographical organization for different movements, and a sensitivity to upper limb impairments and recovery after stroke. The number of muscle synergies did not reflect the level of impairment or recovery in severe chronic stroke patients.These tools measure motor performance and nervous system status, enhancing understanding of stroke recovery.

Corticomuscular coherence (CMC) reflects functional connectivity between the motor cortex and muscles, providing insights into neural control mechanisms. It serves as a potential biomarker for assessing motor function, especially in neurorehabilitation contexts. This study investigates CMC modulation during a standardized robot-assisted reaching task, motivated by the need to understand how assistive technology, like exoskeletons, influences it. 20 healthy subjects performed the task under three conditions: without assistance, and with low and high levels of exoskeleton support. High-density EEG and surface EMG recordings were analyzed to compute CMC in alpha (8–13 Hz) and beta (13–30 Hz) bands. Results indicated no significant differences in CMC across conditions, suggesting that exoskeleton interaction does not alter healthy CMC. These findings support the potential of CMC as a stable measure for motor function in rehabilitation settings, warranting further exploration in clinical populations and with diverse assistive technologies.

In recent years, robotic devices for rehabilitation have gained interest as a possible tool to improve the outcome of the rehabilitative intervention. Nonetheless, objective measures to assess and compare the effects of robotic systems on the motor performance are still missing. In this work a muscle synergy analysis is performed to explore the impact of the upper limb exoskeleton FLOAT on the modular organization of the reaching movement. The results of the study show that muscle synergies are mainly preserved when healthy individuals perform the reaching task with and without FLOAT. This outcome suggests that the use of FLOAT could be beneficial to the recovery of the motor function of the upper limb.

Transcranial direct current stimulation (tDCS) is a potentially effective intervention for stroke rehabilitation. However, conventional tDCS is limited by spatial resolution to specifically target a brain region. Therefore, this study utilized TMS, and computational modeling-guided high-definition tDCS (HD-tDCS). Stroke participants had three visits 1) anodal HD-tDCS stimulation of the primary motor cortex to improve function of the corticospinal tract in the lesioned hemisphere, 2) cathodal stimulation of the dorsal premotor cortex to inhibit use of the cortico-reticulospinal tract in the contralesional hemisphere, and 3) sham. The effect was assessed by qualitative EEG metrics Delta-Alpha Ratio (DAR) and Delta-Theta-Alpha-Beta Ratio (DTABR) as objective outcome measures. Both anodal and cathodal stimulations significantly decreased the DAR and improved Fugl-Meyer Upper Extremity scores. No significant changes in DTABR. Targeted HD-tDCS may improve brain function and reduce post-stroke impairments which could be integrated with robotic based therapy as future work. DAR could be an objective method to assess alteration of brain activity in stroke rehabilitation.

This study aimed to develop a novel neuromuscular coordination-guided rehabilitation exercise through human-machine interaction under isometric conditions and tested its effects on muscular coordination and clinical scores to improve motor function post-stroke. Six post-stroke survivors participated in a muscle synergy-guided exercise as a test group (n = 3) and a force strengthening-guided training as a control group (n = 3). One complete training set consisted of 18 one-hour training sessions for six weeks. Pre- and post-training assessments and one-month and three-month retention tests were performed using a custom-designed robotic device under isometric conditions. While the muscle synergy-guided test group improved intermuscular coordination and Fugl-Meyer Assessment scores, the force strengthening-guided control did not. This pilot study shows that abnormal intermuscular coordination can be malleable to be similar to healthy patterns through human-machine interaction, even for severely impaired survivors, and the therapeutic exercise has the potential to improve motor control in the upper extremity after stroke.

The development of wearable robots offers a promising solution to assist individuals with impaired motor skills by facilitating rehabilitation exercises and aiding in everyday activities. However, the same object is grasped differently depending on what one plans to do with an object next. Therefore, exploring wearable robots that offer broader support for upper limb functions is a valuable research area. In our study, we introduce a wearable robot targeting the hand and forearm, utilizing tendon-driven actuation and a hybrid active-passive mechanism with a spring blade. This device aids in the extension and flexion of the thumb, index, and middle fingers, while also enabling pronation and supination movements of the forearm. This innovation allows users to perform gripping actions in various forearm positions, expanding the scope of rehabilitation training and practical daily activities. Our findings also demonstrate the device’s capability in executing simple everyday tasks.

Upper limb amputation significantly challenges independence, as the human hand is essential for sensing and interacting with the environment. As a consequence, researchers developed complex prosthetic devices, equipped with multiple degrees of freedom, capable of restoring lost functionalities. To enhance prostheses control, current focus is on high-density electromyography (HD-EMG) interfaces, which are susceptible to electrode shift, thus posing reliability issues. To overcome this problem, this study proposes an algorithm to detect and measure the electrode displacement in both lateral and longitudinal directions within an HD-EMG setting. The proposed algorithm achieved a mean accuracy of 79.06% in predicting electrodes shifts, showing potential for real-world application to empower data-driven models over multi-day prostheses usage.

Hand function loss deeply impacts daily activities, highlighting the need for effective rehabilitation strategies. Robot-assisted therapy provides a convenient option for independent rehabilitation, while bilateral therapies enhance motor recovery by leveraging hemispheric interaction and inter-limb coordination. This study introduces a hand robotic rehabilitation platform designed for bilateral therapies using the Leap Motion Controller. The robotic exoskeleton employs an under-actuated linkage mechanism, allowing direct control the motion of the metacarpophalangeal (MCP) joints. In bilateral therapies, the exoskeleton worn on the affected hand replicates the movements of the unaffected hand detected by the Leap Motion Controller. MCP angles are calculated using the dot product of vectors from the metacarpals and proximal phalanges and then used to drive the robot accordingly. A six-week clinical trial was conducted at Centro Hospitalario Benito Menni, involving 13 participants. Results indicated improved performance and ease of control with the robot. Participants found the platform highly usable and effective in motor recovery, experiencing minimal fatigue during sessions.

Stroke rehabilitation plays a vital role in restoring the normal lives of stroke survivors who have impairments due to the stroke. Rehabilitation robots are emerging to replace their traditional counterparts and show promise in providing efficient rehabilitation training. Bio-signal modalities such as electromyography, electroencephalography, or near-infrared spectroscopy, when combined with rehabilitation robots, use the motion intention of their wearers to enhance the effect of rehabilitation therapy. However, many technological advancements are needed in this integration, particularly in understanding the information contained in these signals relevant to robotic control. This paper presents a brain-computer interface that can estimate human motion intention with higher accuracy and improved temporal performance for a hand neuro-rehabilitation robot, SMOVE.

Robotic devices, in combination with virtual reality games, have the potential to increase therapy dosage while enhancing patient’s motivation. Yet, current robotic interventions suffer from poor usability, over-reliance on the availability of trained therapists, and the inability to provide meaningful somatosensory information despite its importance for relearning skillful movements. To address this gap, we co-created two novel haptic rehabilitation robots for in-clinic and in-home rehabilitation capable of high-fidelity haptic rendering during functional reach and grasp training in motivating virtual games together with rehabilitation experts. We evaluated the usability of our solutions with therapists and patients following a mixed-methods approach, gathering quantitative and qualitative data from questionnaires and semi-structured interviews. The results showed good usability and high enjoyment, with the fidelity of virtual object interactions highly praised. Some mechanical design improvements, mainly with regard to comfort, were also identified. Our devices offer naturalistic sensations during training, paving the way for more holistic sensorimotor neurorehabilitation.

After a stroke, stroke survivors often experience severe upper limb impairments that affect their ability to perform daily tasks independently. Current wearable robots usually focus on either hand or wrist assistance, although the combined movements of both are crucial for activities of daily living (ADL). In our study, we introduce a novel wearable robot integrating a soft robotic glove and a wearable wrist robot into a single compact unit. The robot can assist multiple degree of freedom (DOF) movements of both the hand and wrist. The hybrid actuation mechanism, which combines cables and elastic elements, enables natural movement while maintaining high DOF and compactness. Three experiments were conducted with healthy participants to evaluate the robot’s performance and usability. Results indicate promising outcomes in task performance, easy donning/doffing, and user satisfaction. This research highlights the potential of integrating wearable robots in enhancing the independence of stroke survivors.

Individuals with central-nervous-system lesions often face higher fall risks, necessitating improved dynamic balance during walking. At the rehabilitation center of the Ghent University Hospital an instrumented treadmill in a Virtual Reality (GRAIL, Motek) is used for dynamic balance training. An RCT compares GRAIL training to standard balance training, with 28 participants (16 stroke, 6 TBI, 6 SCI) showing promising initial results. Only the GRAIL group demonstrated a significantly improved dynamic balance (CBMS + 8.5, p = 0.002) and balance confidence (ABC questionnaire + 18 points, p = 0.002). Both groups improved walking speed and step length post-intervention; however, solely the GRAIL group showed a decreased double support phase (−6%, p < 0.001), possibly indicating a better dynamic balance.

This study integrates Immersive Virtual Reality (IVR) with Gait Rehabilitation Robotics (GRR) using the Lokomat system and a head-mounted display. The synchronized setup immerses patients in a virtual environment reflecting their physical movements. Fifteen patients with neurological impairments performed tasks combining head movements with lower limb activity while walking in Lokomat. Feedback from satisfaction questionnaires showed strong acceptance and enjoyment of IVR-enhanced sessions. Physiotherapists also saw IVR as a valuable tool for future clinical use, suggesting its integration into standard rehabilitation protocols. These results highlight IVR’s potential in enhancing GRR sessions, prompting further investigation and development for rehabilitation applications.

Neuroscience evidence suggests that personalized, task-specific, high-intensity training is essential for maximizing recovery after acquired brain injury. Robotic devices combined with immersive virtual reality (VR) games, visualized through head-mounted displays (HMDs), can support such intensive training within naturalistic virtual environments with audio-visual stimuli tailored to individual needs. However, the impact of these auditory and visual demands on cognitive load remains an open question. To address this, we conducted an experiment with 22 healthy participants to explore how varying levels of visual, auditory, and cognitive demands affect users’ cognitive load and performance during a shopping task in immersive VR. We found that mental demand had the most significant impact on increasing cognitive load and hampering task performance. Visual demands, although affecting gaze behavior, did not significantly affect cognitive load or performance. Auditory demands showed small effects on cognitive load.

Human physical function is governed by self-efficacy, the belief in one’s motor capacity. In chronic pain patients, this capacity may remain reduced long after the damage causing the pain has been cured. Chronic pain alters body schema, affecting how patients perceive the dimension and pose of their bodies. We exploit this deficit using robotic manipulation technology and augmented sensory stimuli through virtual reality technology. We propose a sensory stimuli manipulation method aimed at modifying body schema to restore lost self-efficacy.

The use of self-paced omnidirectional treadmills equipped with body weight support and virtual reality environments in the clinical practice may allow the training of walking in safe and repeatable conditions across different pathologies. One premise for this to be true is that the conditions of overground and treadmill walking may be similar, so that the latter can be used as an early training to develop the former. Here, we study the equivalence in the spatio-temporal parameters of gait between walking overground and on the Moonwalker, an omnidirectional treadmill equipped with body weight support and virtual reality. Additionally, we show that such gait parameters could be used to distinguish between healthy and parkinsonian patients with similar accuracies for the overground and treadmill condition. Overall, this work contributes to justify the introduction of self-paced omnidirectional treadmill in the clinical and rehabilitative practice.

Rehabilitation robots are frequently used in combination with virtual or augmented reality applications. Impairments in the patient’s visuospatial perception reduces the ability to correctly estimate the distance between virtual objects, leading to a limited task performance. However, the implementation of monocular cues can enhance the performance in patients with impaired visual spatial perception. Since impairments in visuospatial perception are present after stroke, these findings might be crucial when applying virtual or augmented reality-based intervention methods. In order to maximize the therapy outcome, the design of the AR games should be adapted to patients’ visuospatial perception.

The addition of a cognitive task to a balance task impacts postural control. While the effects of dual-tasks have been extensively investigated on the ability of maintaining an upright posture in static conditions or during walking, a limited number of studies focused on other dynamic balance scenarios. In this work, we explored how a motor-cognitive dual-task influenced balance and cognitive performance in unimpaired individuals, in two age groups, under two different dynamic balance conditions. Subjects stood on a moving robotic platform that tilted either in response to subjects’ weight shift or based on a preprogrammed motion trajectory. Subjects performed two versions of the Stroop Test: the classic format and a digital version. As expected balance performance always worsened when adding the cognitive task. Conversely, cognitive performance deteriorated with cognitive task complexity and was also influenced by the difficulty of motor tasks.

This study presents an approach to real-time data acquisition and processing using Inertial Measurement Units (IMUs) in a System-on-Chip (SoC) based on Field-Programmable Gate Array (FPGA). By leveraging the parallel processing capabilities of FPGAs, the system achieves a low latency, essential for real-time applications where accurate orientation estimation, such as in neurorehabilitation. In this context, the use of IMUs and FPGAs can be particularly beneficial for movement assessment, biofeedback, and the control of assistive devices, which are fundamental in rehabilitation. The experimental results demonstrate the effectiveness of this approach, with minimal latency, which is critical for therapies that depend on the patient’s rapid response to stimuli. This paper discusses the materials and methods used, including the SoC architecture for data processing, showing the potential of FPGA-based SoCs to enhance performance in rehabilitation-focused applications, offering significant improvements in the precision and efficiency of monitoring and control systems in this field.

The management of a large neuromechanics facility is a complex task, that requires the selection of the appropriate software to control system operations, activate various subsystems, integrate sensory data, contextualise it temporally and geometrically, and store all relevant information. This work proposes the use of ROS as a base control and data collection system for large-scale neuromechanics laboratories. In the proposed solution, ROS and ROS 2 work together: the first one being responsible for data transmission and the second one for synchronisation through a trigger card. In addition, mass storage of the obtained data is addressed. To test the capabilities of the system, tests have been carried out simulating the normal operation of the installation using 3 test benches of the Eurobench project, which include many of the most commonly used sensors in a laboratory of this type such as IMU, force platforms, load cells and robotic systems.

This paper presents insights into the evolution of measurements in neurorehabilitation from the perspective of a 30-year professional experience. The author acknowledges a number of positive advancements but also highlights the risks of potential misdirection.

Assist-as-needed approaches have shown to be effective in robotic stroke rehabilitation. We here present an assist-as-needed approach for MirA, an upper-limb robotic rehabilitation method based on mirror therapy. We use a bi-manual rehabilitation exoskeleton (ALEx-RS, Wearable robotics) to assist the hemiparetic arm during virtual reality tasks. Our goal is to adapt the assistance based on the movement accuracy of the healthy arm and the participant’s performance. We present data of the exoskeleton’s adaptive assistance in a proof-of-concept test with one healthy subject, demonstrating that the system decreases assistance when the subject performs well and increases it when performance declines.

Physics-based computer simulations are a powerful tool to define personalized robotic rehabilitation treatment planning for individuals with spinal cord injury (SCI). Current models often lack detailed representation of upper body muscle forces, which are crucial as these subjects often rely on walkers for mobility. This work investigates if adding physiological muscle torque actuators improve the prediction of upper body mechanics compared to having ideal torques. We calibrated muscle torque-generating functions for a 3D full body model using dynamometry measurements. An ideal-torque-driven and muscle-torque-driven model were tested in unassisted and walker-assisted gait predictions with different configurations. Results suggest that incorporating muscle torque actuators improve the physiological accuracy of simulations while not reducing significantly computational efficiency. Future work will focus on implementing calibrated lower body muscle models in gait predictive simulations to support clinical decisions.

Patients with neurological disorders often have upper extremity impairments. This work aimed to develop a rehabilitation device which enables neuromuscular training with adjustable force resistance and position assistance. The rehabilitation device consists of two compact cable-driven units (CDUs). Each CDU can easily be fixed to a suitable support surface. Using PID force control and impedance control algorithms, the device can produce user-defined resistive training with position assistance. A user interface is developed to setup the training intensities and control parameters. A single-case test for technical evaluation showed that the force control algorithms produced a mean force error of 3.78 N during static position, and 2.76 N during isotonic training. The impedance controller yielded a position tracking error of 0.08 m. It was concluded that the rehabilitation device managed to offer neuromuscular training with satisfactory accuracies.

Hand impairments following a stroke can impact their ability to perform daily activities including grasping, reaching and hand manipulation. Therefore, there is a need for more effective motor rehabilitation interventions post-stroke. Brain-Computer Interface (BCI) systems establish direct communication between the brain and the external environment, facilitating the repeated activation of closed-loop motor circuits damaged by stroke. Detecting a user’s hand movement intentions in EEG-based BCI systems remains one of the most challenging tasks within the BCI field. This study introduces a BCI-based dexterous hand rehabilitation robot for post-stroke grasping training, combining an EEG-vision hybrid intention recognition system with an 8-degree-of-freedom (DOF) hand rehabilitation glove.

Providing arm weight support increases range of motion and breaks abnormal movement patterns such as flexion synergy. A simple robot with one actuated DOF could be ideal for proving arm weight support and enhance feedback in home settings. This article briefly presents the kinematics and optimization of such a robot– MARS (Motorized Arm Support). It is common to add redundant degrees of freedom (DOF) to increase the robot’s workspace. Thus, two kinematic chains of MARS are considered, 4-DOF MARS and 5-DOF MARS with one redundant DOF for the optimization problem. The 5-DOF MARS reaches 4.4% more work area of the arm but increases the complexity of the system as it requires additional mechanical elements such as dampers to prevent impact forces at the arm. Thus, to keep the system simple, the small loss in work area is considered acceptable and the 4-DOF kinematic chain of MARS is chosen for further development.

An instrumented chair (IC) is proposed in this study to evaluate the sitting posture in a simple manner. This chair is capable of measuring the forces and pressure distribution during sitting in critical regions, i.e., under feet, sitting, and armrest regions, which are sensorized. One of the authors of this manuscript participated in these experiments. The results show that the difference in forces at the right and left side and also the pressure distribution at the hands and feet separately during sitting and standing using this relatively low-cost and accessible system.

In stroke rehabilitation, tailoring assistance to individual needs is crucial for more effective training. This study investigates the control architecture of an artificial partner (AP) inspired by game theory. The AP modulates assistance in a planar reaching task using adaptive control strategies. We compare AP performance in “lazy” and “generous” conditions. Results show that the AP adjusts its assistance effectively based on game-theoretic principles. This approach shows promise for enhancing robot-assisted rehabilitation through personalized therapy. Future research will explore the long-term effects of these policies and refine the AP’s sensory system and state observer for improved precision.

Hybrid exoskeletons combine functional electrical stimulation (FES) and robotic assistance to restore and rehabilitate movement. Real-time control of FES parameters is necessary to maximize outcomes. Recently, muscle deformation captured by ultrasound imaging has been demonstrated as an effective human machine interface. Here, we examine if ultrasound imaging can track joint kinematics in the presence and absence of FES to assess its viability for real-time control of stimulation parameters. Muscle deformation from ultrasound was correlated to limb movement using a simple image similarity metric. One healthy participant performed volitional and FES-driven wrist, ankle, and knee flexion/extension while simultaneous motion capture, surface electromyography, and ultrasound were collected. Ultrasound imaging tracked joint kinematics with and without FES with mean errors of 1.7 ± 20.7% and 3.9 ± 18.1%, respectively, averaged across all three joints. Our results demonstrate the feasibility of ultrasound to track stimulation driven movement and potentially adjust FES parameters in real-time.

In this study, we propose a hybrid gait orthosis that combines the advantages of functional electrical stimulation (FES) and a gait orthosis. The hybrid gait orthosis uses pneumatic muscles to control the hip joint and an electrical stimulation system to control the knee joint. It is intended to overcome the limitations of existing powered gait orthosis systems, which have difficulty controlling the knee joint. The hybrid gait orthosis was tested on a paraplegic patient, and the results showed that the patient was able to walk with a natural gait. The hybrid gait orthosis has the potential to improve the quality of life of paraplegic patients by allowing them to walk more naturally and with less effort.

Exoskeleton-assisted rehabilitation delivers consistent and high levels of training intensity as an innovative therapeutic option for the restoration of functional gait in stroke survivors. Evidence on neurophysiological response induced by this training in end-users is limited. Understanding cortical and muscular activity can be used for user adaptation during training. This work aims to describe the data processing pipeline for Event-Related Spectral Perturbation (ERSP) and Cortico-Muscular Coherence (CMC). Variations in the ERSP suggest that modulation of neural activity does not occur during the Exo session.

Rehabilitation following Total Knee Arthroplasty (TKA) is crucial for patient outcomes. With the ageing population, the demand for such surgeries is rising, necessitating advancements in patient care and reducing healthcare burdens. Robotics offers a promising solution, with various systems emerging.This paper introduces a robotic arm-based (KUKA LBR Med 14) rehabilitation system with real-time anatomical tracking. Using bone surrogates (Sawbones), the system was tested for passive Flexion-Extension (FE) and Ab-Adduction (AA) movements of the shank. The robotic arm’s End-Effector (EE) was attached to the shank with a custom elastic band system, keeping the thigh nearly stationary.A key feature is real-time tracking, allowing the robot to adapt its motion to the femur’s position, ensuring precise control over femur-tibia rotational displacement. Promising results showed effective path tracking for passive motion tasks, aided by an impedance controller for secondary knee joint movements. Future work includes developing a non-invasive tracking system and active rehabilitation exercises.

The design and user interaction of walking assistance devices are crucial for people with gait impairments. This study focuses on optimizing anchor point placement in a quasi-passive exosuit for hip flexion assistance to lower metabolic cost during walking. Simulations in OpenSim explored different actuator positions to minimize metabolic consumption. Simulation results indicated that anchor points on the frontal part of the knee reduced metabolic cost most effectively. However, experimental results from 11 healthy subjects walking on a treadmill showed that wearing the exosuit increased metabolic consumption, especially at higher speeds. These findings suggest that while simulations can guide design optimizations, they require experimental validation to ensure practical effectiveness.

Locomotion-focused neural rehabilitation therapies intensively practice gait motion on treadmills or overground for patients with neurological injuries. Traditionally, skilled therapists or robotic devices are required to assist with both the patient’s body weight support and pelvis movement, making the rehabilitation process both physically demanding and economically burdensome. In this abstract, we briefly summarized our developed detachable body weight support (BWS) rollator equipped with wearable sensors to facilitate overground gait training, aiming to enhance the user experience with minimal motion compensation. The developed BWS rollator system integrates wearable sensors, including instrumented shoes and harnesses, to capture the user’s gait performance in real-time. Various control strategies were developed to assist individuals with different levels of mobility difficulties. The system’s usability was validated through physical experiments with able-bodied subjects. The developed detachable BWS systems and control strategies are expected to be utilized in neural rehabilitation. Future research should focus on long-term clinical trials involving subjects with neurological disabilities.

The Sit-To-Stand (STS) transition is a critical activity for maintaining physical fitness and independence among elderly individuals. Declining STS ability due to agerelated factors can lead to dependence and increased fall risk. This paper introduces the SkyWalker, a robotic rollator designed to provide adaptive STS trajectory motion support using advanced 3D visual pose estimation. Equipped with depth cameras and machine learning algorithms, the SkyWalker detects and responds to user movements in real-time, ensuring safe and smooth transitions from sitting to standing. Our vision- based approach employs the Google Mediapipe Human Skeleton algorithm to accurately capture user posture and classify STS phases, enabling the robot to dynamically adjust its STS trajectory based on user movements. Experimental results with seven participants demonstrate the effectiveness of our adaptive STS trajectory system in enhancing user stability and comfort compared to predefined trajectories. The SkyWalker’s adaptive assistance significantly improves synchronization of the robot movement with the user’s natural movements, reducing fall risk and promoting independence.

Difficulties in standing up and sitting down can greatly impact the quality of life of older adults. Robotic rollators equipped with moving handles can provide support to prolong the ability to live independently. However, appropriate trajectories are still unknown while the corresponding human-machine interaction dynamics are poorly understood. This study proposes bio-inspired, easy-to-implement assistance trajectories and investigates their biomechanical effects on standing-up and sitting-down motion in 15 young adults using moving handles of a robotic assistance device in a motion capture lab. The results show that moving-handles assistance can provide up to twice as much vertical support as fixed handles while potentially reducing lower and upper limb muscle demands, and reducing the destabilizing effect of horizontal forces on the handles. Furthermore, participants’ perception of provided help correlates well with the magnitude of the vertical force support. The study’s subsequent phase comprises the ongoing evaluation with frail older adults.

This study investigates the effectiveness of a multimodal feedback interface for assisted navigation using a smart walker, addressing mobility challenges faced by older adults. The methodology involved equipping a walker with sensors and feedback components, and testing it with 12 healthy participants using visual, auditory, and haptic feedback modes in a controlled environment. Results indicated significant differences in user performance across feedback types, with vibration feedback being more comfortable and less stressful compared to others. The findings highlight user preferences for combined feedback modalities, contributing to the design of improved assistive navigation systems for older adults.

Patients with neurological disorders often have gait impairments. This work aimed to develop a computer-vision-based mobile gait trainer which supports users during free overground walking. The mobile gait trainer consists of three submodules: a Mecanum-wheel-driven frame with a body-weight-support (BWS) mechanism, a pose estimation module and a motor control system. Secured by the BWS, the user walked overground at preferred speeds. The webcam on the top frame estimated the shoulder movement based on the MMPose algorithms. Kinematic analysis yielded the target speeds for the trainer to follow the user. The motor control algorithms enabled the BWS to relieve the target load and the trainer to move. Preliminary test showed that the BWS mechanism produced a mean force control error of 2.03% for free walking, and 4.57% during obstacle climbing. The trainer followed the user with a speed error of 0.17 m/s. It was concluded that the trainer managed to support free overground walking.

For patients with physical or cognitive impairments, there is a pressing need to improve the recovery of their physical functions through specific therapeutic training processes. Thanks to its customization potential, Rehabilitation Robotics is becoming an essential area of Robotics for Medicine and Healthcare. This work concerns the design of a human-robot interaction module for gravity compensation to be used in robot-assisted rehabilitation of the hand and upper limb. The proposed module is designed to be connected, on one side, to a collaborative manipulator as an end-effector and, on the other side, to the patient’s forearm. The aim is to compensate for upper limb gravity during rehabilitation exercises while ensuring complete freedom of movement in the three-dimensional operating space. The perspective is to use the proposed solution as an enabling technology for more in-depth and targeted rehabilitation sessions once coupled with proper user-centred control strategies for the manipulator.

Wearable robotic exoskeletons are currently the only option to restore upright mobility in persons with spinal cord injury (SCI). However, such weight-bearing locomotion exposes this population to a greater risk of bone fracture during exoskeletal-assisted walking (EAW). The increasing popularity of robotic exoskeletons emphasizes the need for the quantification of human-robot dynamics and joint loading during EAW. Thus, the goals of this study were to quantify 1) human-robot dynamics and 2) joint forces at lower-limb joints of persons with SCI during EAW in three FDA-approved devices. Our findings indicate that persons with SCI experience similar joint forces across the three FDA-approved exoskeletons. Joint dynamics and loading during EAW differ substantially from unassisted walking. Our study provides a computational framework to quantify the forces exerted on the long bones of persons with SCI during EAW. This work provides the foundation for future evaluation of fracture risk during EAW.

In this study, we introduce a novel deep reinforcement learning (DRL) based approach for controlling lower limb rehabilitation exoskeletons (LLREs). Our method employs a neural network-based controller that accurately forecasts real-time commands for the exoskeleton’s actuators using only proprioceptive signals from the LLRE. This controller is trained within a sophisticated virtual simulation environment integrating a comprehensive human musculoskeletal model and an exoskeleton interaction model. To enhance adaptability, we utilized domain randomization during training to simulate diverse patient musculoskeletal conditions. We validate the effectiveness and robustness of our DRL-based LLRE controller across various neuromuscular conditions during walking, evaluating key metrics such as stability and gait symmetry. This innovative approach supports seamless deployment of trained controllers onto physical hardware through sim-to-real transfer, eliminating the need for patient-specific experimentation and parameter tuning. Our work represents a significant advancement in LLRE control methodology, promising enhanced functionality and adaptability for real-world applications.

Neurological diseases impact around 1 billion people globally, with the numbers expected to increase due to population growth and aging, leading to a decline in lifestyle quality. The need of the hour is to develop a comprehensive methodology for ankle, balance, and step-taking rehabilitation for improving postural adjustment strategies (motor learning) in patients with different disease severity. To meet this need, we propose the Integrated Balance Rehabilitation (I-BAR) framework, designed to boost rehabilitation effectiveness through objective assessment, personalized therapy, and adaptability to different disability levels. I-BaR combines ankle-foot preparation, balance, and stepping phases into a cohesive process. It offers multi-modal feedback: visual feedback via virtual reality, vestibular feedback through robotic platform-induced perturbations, and proprioceptive feedback using haptic feedback, enabling patients to improve their balance abilities comprehensively.

This study investigates the static and dynamic responses of a new sensor prototype, currently under development, for force measurement. Constructed using carbon nanotubes (CNT) dispersed in an optimised elastomeric matrix, these sensors possess material characteristics that make them suitable for footstep detection. The paper presents the characterisation of a sample and offers a comparative analysis with Force Sensitive Resistors (FSRs), which are among the most commonly used solutions for detecting force changes in footstep applications, particularly in exoskeletons. An evaluation of their performance is provided, suggesting their potential application for this purpose.

Stroke survivors improve their motor functions through two main mechanisms: reduction of the impairments and development of compensatory strategies. To disentangle these mechanisms and to tailor assistance to individual needs, detailed descriptions of the patients are crucial. Here we describe an experimental apparatus that includes a set of inertial measurement units (IMUs) and a planar manipulandum, to be used as assessment and rehabilitation tool. We propose a pipeline for generating personalized musculoskeletal models to reconstruct whole-body movements, with the specific aim of distinguishing between compensation and true recovery for tailored rehabilitation interventions.

Following spinal cord injury (SCI), upper extremity (UE) paralysis or weakness impede one's ability to carry out activities of daily living (ADLs). Such a limitation drastically lowers a person's level of independence. The main objective of this study was to assess the benefits of an UE myoelectric-powered wearable orthosis (MPWO) produced by MyoMo, Inc. (Boston, MA) on improving UE motor function in order to enhance ADLs and quality of life in individuals with subacute SCI. These preliminary findings imply that UE-MPWO device-assisted rehabilitation may increase participants’ UE activities, leading to improved function.

Rehabilitation exoskeletons face acceptability issues due to their inability to produce trajectories representing therapists’ desires. Learning by Demonstration can offer a solution to this issue by using therapists’ demonstrations to teach the robot how to perform movements. While various methods for teaching movements to a robot from human demonstrations and manipulation have been proposed, no discussion about the best one for rehabilitation purposes has been conducted. Our work aims to discuss a Hidden-Markov-Model-based (HMM) approach, integrated with a human-likeness and a joint-synchronization algorithm, and compare it with two other well-known approaches we previously validated for rehabilitation. Applying the three methods to some therapists-generated datasets, produced with the exoskeleton AGREE and representing 5 different exercises, we verified that our HMM approach is better both in terms of human likeness (as measured by the SPARC metric), and compliance with therapy time requirements (needing less than a second of computation).

This study compares knee kinematics and dynamics obtained by the NanoStim and optical motion capture systems. The NanoStim is a wearable device that delivers electrical stimulation and collects biofeedback data. Initial results from four participants show no significant differences between motion capture and the wearable. Future research will validate NanoStim feedback and develop optimized stimulation patterns for muscle rehabilitation in the elderly.

Upper Motor Neuron Syndrome (UMNS) is associated with impairments of upper and lower limbs functions that can severely affect patients’ independence in daily activities.

Focal Muscle Vibration (FMV) is receiving increasing attention as a tool for neuromuscular rehabilitation, particularly for stroke and spinal injury rehabilitation. Frequency characterization is a key factor in maximizing FMV related rehabilitation outcomes. We present a well characterized design of FMV device that can be used for various research applications. We also present an experimental scheme for characterizing the vibrational behavior of the device. Finally, we design a 3D printing casing to encapsulate the device and an electronic circuitry to control the vibration behavior of the device. The outcome of the project was a low-cost, user-friendly, wearable FMV device with a range of frequency settings specifically suited for neurorehabilitation.

Ankle dorsiflexion is a fundamental component for lower limb functionality and the quality of life of an individual that involves the coordinated action between several muscles and nerves. This movement can be affected in patients with various neurological pathologies of both central and peripheral nervous systems, causing alterations in sensory and/or motor function and whose physical impact can vary according to the severity, location, and extent of each patient's neurological injury. This study investigates the effect on selective ankle dorsiflexion and tibialis anterior muscle activation during the transition from sitting to standing in patients with various neurological conditions that have resulted in a loss of the described function, through an intervention protocol combining the application of a functional electrical stimulation (FES) device and invasive electrical stimulation on tibialis anterior musculature in an echo-guided manner.

This study integrates synergy-based Functional Electrical Stimulation (FES) with a robotic exoskeleton for upper limb rehabilitation, aiming to improve muscle recruitment, delay muscle fatigue, and reduce motor energy consumption. Preliminary tests with 3 non-disabled subjects indicate that the hybrid system can reduce energy consumption while preserving a good trajectory tracking. Future studies should evaluate the efficacy of this hybrid approach in enhancing rehabilitation outcomes for neurological patients.

Recent advances in orthotic devices incorporate active components aimed at enhancing human functional performance by either delivering or absorbing energy. Functional Electrical Stimulation (FES) via surface electrodes represents a significant development in this field, offering substantial therapeutic potential. However, a primary challenge lies in the precise positioning of these electrodes, which is crucial for optimizing therapeutic outcomes. In order to address this challenge, a 3D-printed orthotic device capable of accurately positioning electrodes on the forearm is proposed. To achieve customization according to the patient’s unique anatomical features and optimal electrode positions, a comprehensive arm scan precedes the design and fabrication of the orthoses. The design procedure employs commercial software and is automatable. Additive manufacturing techniques are well-suited for the production of the resulting complex geometries. Preliminary results indicate that the approach is viable and that the solution has the potential to increase accessibility to the technology and patient independence.

Rehabilitation interventions that employ brain-computer interfaces (BCIs) require numerous sessions spanning over weeks or even months. Motor imagery (MI) based BCI models need to be calibrated at the start of each session with up to 20–30 min of data to achieve acceptable performance. Minimizing calibration times is important to increase the effective therapy time. Transfer learning could be used to address this issue, however, its performance with multi-session data in neurorehabilitation settings remains relatively underexplored. Here, transfer learning was applied to MI data from 1 naïve individual with SCI, which was combined with previous sessions data to reduce the calibration time needed at the beginning of each session. The use of transfer learning resulted in classification accuracy similar to standard calibration, but significantly reduced the amount of data required to be collected, from 18.06 ± 0.6 min to 3.84 ± 0.01 min (p < 0.01). Future research further improve transfer learning approaches to increase BCI classification accuracy.

Compared to surface electromyography, force myography, or FMG, setups can be very cost-effective and robust, and they offer the advantage of not requiring skin contact. FMG as a modality is inherently correlated with muscular force, and should by extension be more directly correlated with measurable net force output in healthy subjects as well. In this study, we set out to prove this hypothesis by comparing the performance of a regression algorithm predicting isometric finger pressure based on FMG, as opposed to electromyography. Additionally, we demonstrated the effectiveness of a low-cost system, which concurrently with the robustness of FMG sensors, would make them an attractive solution to be used in resource-constrained settings, thus fostering a more widespread adoption of electronic prostheses in low- and middle income countries.

Gait rehabilitation often faces difficulties translating into daily life once patients leave the clinical setting. Wearable systems that trigger individualized movement reminders can promote awareness of gait parameters. This study presents a modular framework for developing personalized real-time biofeedback systems in gait rehabilitation, demonstrated in a case study where knee range of motion feedback was provided via smartwatch. The device is currently being piloted in real-world use, with frequency and distribution of alerts under evaluation.

Stroke and Parkinson’s disease often result in significant gait impairments reducing mobility, independence and overall quality of life. While biofeedback has proven effective in restoring healthy gait, it is typically designed without considering the patient perspective. Therefore, this study aimed to implement a co-design process for personalized real-time biofeedback, to determine its impact on patient motivation and gait outcomes. Five participants (4 stroke, 1 stroke and Parkinson’s disease) participated in multiple personalized real-time biofeedback-based gait training sessions using a Computer Assisted Rehabilitation Environment, actively shaping the features (e.g. visual biofeedback presentation, background settings, sound preferences, etc.). Continuous adjustments to the biofeedback and its difficulty effectively maintained patient motivation and engagement. Combined with eliciting changes in gait outcomes, personalized real-time biofeedback represents a promising approach to improve current gait training approaches, with potential implications in the rehabilitation process.

Stroke and Parkinson’s disease (PD) are prevalent neurological conditions causing significant disability in the elderly. Gait impairments are common targets for rehabilitation, yet responses to interventions vary widely. Within the scope of the StimuLOOP project, this monocentric, proof-of-concept clinical trial, is conducted to explore the efficacy of personalized real-time biofeedback for enhancing gait quality. Participants aged ≥18 years with stroke (≥30 days post-event, Functional Ambulation Category ≥3) or PD (UPDRS III gait/postural instability scores 1–3) undergo 15 days of personalized real-time biofeedback gait training. The dominant gait deficit of the participant is objectively evaluated in order to tailor the biofeedback by selecting a target parameter that the participant should actively modulate during training. Preliminary findings suggest participants are able to use personalized real-time biofeedback to adjust their dominant gait deficit, and subsequently maintain their gait adaptations in the absence of visual biofeedback. Ongoing analyses aim to validate these observations and expand our understanding of how motor learning evolves within and across these training sessions.

Real-time biofeedback, particularly visual, is gaining attention in neurorehabilitation, especially for gait training. However, its impact on cognitive load remains poorly understood, despite its role in influencing gait performance. We present here the protocol for a study where 40 healthy participants will be recruited and asked to complete a series of gait tasks with and without visual real-time biofeedback. This work aims to investigate the cognitive load associated with different visual real-time biofeedback paradigms during gait training. Cognitive load will be evaluated using electroencephalography of the frontal and parietal lobes, power spectral density analysis of the corresponding channels, and a questionnaire. The findings will enhance our understanding of how visual biofeedback affects cognitive load, informing future designs to optimize both cognitive load and gait performance in rehabilitation settings.

Repetitive transcranial magnetic stimulation (rTMS) has been extensively used to modulate cortical excitability and treat neurological disorders. However, the extent to which rTMS is effectively transmitted to spinal motor neuron output remains unexplored. In this study, we investigated the corticospinal transmission of various rTMS frequencies to motor unit activity. Eight participants performed low-level isometric contractions of thumb flexion while rTMS pulses were delivered to the primary motor cortex at 70% of the resting motor threshold. Stimulation frequencies of 5, 10, 20, 30 and 50 Hz were used. High-density surface electromyograms recorded from the thenar muscles were decomposed into motor unit spike trains, and the z-coherence between the rTMS stimuli (input) and the cumulative spike train (output) was calculated. Significant input-output coupling was observed at frequencies of 20 Hz and higher. These preliminary findings suggest that the efficiency of the transmission of the synaptic input generated by rTMS to the neural drive to the muscle is frequency dependent.

The learning of a new motor task entails neural changes in spinal pathways. This study aimed to investigate alterations in the common synaptic inputs to spinal motor neurons during the short-term acquisition of a force-matching skill. 11, 13 and 17 participants had the first dorsal interosseus, tibialis anterior, and vastii muscles tested, respectively. They performed 15 isometric force-matching contractions at 5 or 10% of the maximal voluntary contraction. High-density surface electromyograms were decomposed for two selected trials: pre- and post-learning. The motor units were matched between trials and the coherence between cumulative spike trains was calculated within alpha (5–15 Hz) and beta (15–35 Hz) bands. In both individual and synergistic muscles, improvements in force-matching were followed by reductions in the alpha band, but not beta band. These findings suggest that during the learning of a new task, the central nervous system filters tremor noise oscillations unrelated to the required task.

Muscle specific rotational tuning curves are constructed to describe human standing balance. Task-dependent changes in the preferred direction of these tuning curves demonstrate the neural drive to muscles is independently sculpted to each muscle in healthy young adults. Changes to this flexible control scheme my underlie impairments in standing balance.

To investigate how changes in ankle position affect shared neural inputs across triceps surae muscles during ankle plantarflexions, high density surface electromyography was used to measure motor units in the medial and lateral compartments of the soleus (SM, SL) and gastrocnemius (GM, GL) in sixteen healthy adults. Participants performed isometric plantarflexions at 20% maximal voluntary contraction (MVC) at three ankle positions: 20° plantarflexion (PF), neutral (PF0°), and 20° dorsiflexion (DF). Intermuscular coherence of GM-GL and SM-SL in the delta, alpha and beta bands were analyzed. Results showed that force fluctuations decreased with greater ankle dorsiflexion. Regardless of the muscle pair, alpha band coherence was highest at PF20°, and beta band coherence was lowest at the neutral position. Delta-band coherence between GM-GL was consistently higher than that of SM-SL at all positions. Changing ankle positions impacts shared neural inputs in the alpha and beta bands across triceps surae muscles.

The purpose of this study was to quantify the effects of 12 sessions of ischemic conditioning (IC) on mean discharge rate and estimated common synaptic drive (CSD) during a fatiguing contraction in chronic stroke survivors. On average, following IC we found an increase in task duration and decreased oscillations in CSD at the beginning of the task. The results suggest that an IC intervention can improve aspects of rate coding during a fatiguing task and that IC could potentially be used as an adjunct to rehabilitation to optimize neuromuscular fatigability post stroke.

To enhance communication aids driven by code-modulated visual evoked potentials (cVEP), selecting the right language model (LM) is paramount. With the new release of GPT4 turbo, this study compares “gpt-3.5-turbo-0125” (GPT-3.5T) and “gpt-4-turbo-2024-04-09” (GPT-4T). Given the increased time per token for GPT-4T compared to GPT-3.5T, the hypothesis posits that the operational characteristics of GPT-3.5T may better suit the unique demands of cVEP communication aids. By examining parameters such as user experience during BCI spelling in terms of Output Characters per Minute (OCM), accuracy and cost per session, this study aims to identify the most effective LM model for advancing cVEP-based communication aids.

Traditional rapid serial visual presentation (RSVP) spellers using brain-computer interface (BCI) technology usually employ a set of 27 alphabetic characters, with at least three control symbols, presented sequentially for user selection. We propose that only a reduced subset of 12 characters/symbols be presented at a time for selection. The characters in this reduced set will be those most likely to follow the user’s previous selections based on a generic corpus. This approach is expected to reduce the time required to select the desired character and, therefore, increase the efficiency of the system compared to traditional RSVP spellers. The system was tested on one healthy subject, who achieved an accuracy of 100% and an average typing speed of 1.9 characters/min. While these results did not meet initial expectations regarding typing times, future modifications are proposed to continue improving the efficiency of these systems.

Code-modulated visual evoked potentials (c-VEP) are a promising control signal for non-invasive brain-computer interfaces (BCI). Previous studies showed that this paradigm can reach higher performance and reliability than previous approaches, such as P300 potentials or steady-state visual evoked potentials. However, a calibration phase is still needed before starting to use the system, which limits the applicability of the technology. In this exploratory study, our goal is to eliminate the need for this calibration phase to improve the usability of c-VEP-based BCIs. To this end, we designed, developed, and tested a new bitwise reconstruction strategy that takes advantage of deep learning and transfer learning to overcome inter-subject variability. We tested the method on a database of 10 subjects that included recordings with binary (i.e., black and white) and non-binary (i.e., 5 shades of gray) stimulation paradigms to reduce eyestrain. Results show that the proposed method reaches a maximum average accuracy of 95% using 9 stimulation cycles without calibration. These results represent a promising step forward towards more practical c-VEP-based BCIs.

Despite the excellent performance of brain-computer interfaces (BCIs) based on code-modulated visual evoked potentials (c-VEP), these systems are inherently synchronous, resulting in constant selections even when the user is not paying attention to the visual stimuli. This significantly limits their application in real-world environments. The objective of this pilot study is to propose a multi-window correlation-based approach to implement both asynchronous and early stopping stages in real-time. In our system, a decision is made only when the system is confident that (1) the user is receiving visual stimuli and (2) the user is paying attention to a specific command. This approach was assessed offline with eight healthy users, achieving 100% accuracy in asynchronous detection and a mean decoding accuracy of approximately 95% with only 1.8 s per trial. These results demonstrate the effectiveness of the proposed method in monitoring user attention while maintaining competitive performance in c-VEP-based BCIs.

People with Tetraplegia will typically require costly, round-the-clock care over the course of their lives due to lack of motor function. Brain computer interfaces (BCIs) are capable of identifying imagined motor intention for individual limbs; however, high-cost electroencephalography (EEG) devices are typically used. In this paper, a comparison of 4 published motor imagery (MI) BCIs is undertaken. BCI Competition IV Dataset 2b is used to benchmark the four and the relatively low-cost OpenBCI EEG cap is then used with each, in a custom experiment. Results obtained from the OpenBCI EEG cap suggest that it is suitable for MI tasks, though this should be confirmed with a larger sample. Also, 2 of the 4 MI BCI’s stand out for decoding accuracy, though all performed well.

This preliminary study investigates the reliability of power-based EEG metrics, particularly within the beta frequency band, in stroke patients. Seven stroke patients and nine healthy controls participated in various motor tasks aimed at assessing longitudinal neurophysiological changes during rehabilitation. The primary goal was to identify consistent power-based EEG measures, including event-related desynchronization (ERD) and post-movement beta rebound (PMBR), that can reliably track recovery. By establishing reliable metrics, this study aims to improve the precision of EEG-based monitoring tools, facilitating more personalized and effective rehabilitation protocols for stroke patients. These findings may contribute to the development of robust neurophysiological markers critical to optimizing rehabilitation strategies.

Robot-assisted therapy contingent on the intention to move ensures the active engagement of patients during training. However, detecting the intention to move in severely impaired patients with no visible movement is a challenge where physiological signals such as Electromyogram signals (EMG) can be used. An effective EMG-driven robot-assisted therapy in severely impaired patients should provide naturalistic human-machine interaction, which requires an optimal EMG detector with high detection accuracy and low latency. Non-availability of ground truth about the presence/absence of EMG in severely impaired patients with no movement is a challenge, which hinders the computation of detection latency and accuracy. Therefore, this paper identifies an optimal EMG detector without the ground truth about the presence of residual EMG signals. An unsupervised approach using total variation distance was used for this purpose to distinguish between the rest state when the muscle is fully relaxed and the move state where there could be muscle activity. The analysis was done on residual EMG data from one severely impaired stroke patient. The results reveal that the modified Hodges and approximate generalized likelihood ratio (AGLR) detectors maximally separate the rest and move states. The AGLR detector showed poor performance in both detection latency and accuracy, whereas the modified Hodges detector demonstrated better performance, making it a potentially better choice for personalized EMG-driven robot-assisted therapy.

Closed-loop brain-machine interfaces (BMIs) hold significant promise for restoring autonomy to motor-disabled subjects, such as amputees, tetraplegic patients, and others with motor impairments. Current research in BMIs focuses on restoring proprioceptive feedback through direct cortical stimulations. Typically, these systems are tested using animal models, such as mice, to evaluate their real-world applicability. However, few studies have explored informing these systems with in-silico models to predict proprioceptive feedback. In this study, we take a first step in this direction by investigating whether a musculoskeletal forelimb model of the mouse can be controlled to achieve reaching movements. We demonstrate that this musculoskeletal forelimb model can reproduce reaching movements in good agreement with recorded experimental data from mice. In the future, this model is intended to serve as the basis for artificial cortical feedback by predicting proprioceptive signals from the computational muscles in the model. We hypothesize that using this proprioceptive feedback will help to develop enhanced closed-loop neural prostheses.

This paper presents the development of a Brain-Computer Interface (BCI) as a neuromodulation system that detects cortical patterns related to the planning and execution of upper limb movements and closes the motor rehabilitation loop through Functional Electrical Stimulation (FES) with the H-GAIT neuroprosthesis. The foundational elements of the proposed system are biopotentials such as electroencephalography (EEG) and electromyography (EMG), cortical patterns, spatial filters, and pattern detection algorithms. Offline and online biosignal capture protocols were developed to first characterize the selected cortical pattern, then detect it online, and finally close the BCI-FES loop, resulting in a low-latency system capable of inducing neuroplasticity.

Transcranial static magnetic field stimulation (tSMS) is a recent non-invasive brain stimulation (NIBS) technique that uses static magnetic fields to modulate neuronal activity. tSMS is safe, portable, and user-friendly, capable of reducing corticospinal excitability with effects lasting up to 30 min post-stimulation. Research has shown potential clinical applications, but further studies are needed to optimize protocols and fully realize tSMS's therapeutic potential.

This study explores the potential of non-invasive brain stimulation, specifically static magnetic field stimulation (tSMS), to modulate the cortical excitability of lower limb muscles. Justification for this research stems from the successful use of tSMS in neurorehabilitation for conditions like stroke, Parkinson’s disease, and ALS, primarily targeting the primary motor cortex (M1) for upper limb rehabilitation. However, many neurological conditions, such as spinal cord injury (SCI), predominantly affect lower limb motor function. Recent findings suggest that the cortical representation of lower limbs is more superficial than previously thought. This pilot study aims to demonstrate that applying tSMS over the cortical area representing the lower limbs (M1INF) for 30 min can modulate cortical excitability, measured via motor evoked potentials (MEPs) elicited by transcranial magnetic stimulation (TMS).

Non-invasive brain stimulation (NIBS) techniques, including transcranial magnetic stimulation (TMS), transcranial direct current stimulation (tDCS), and transcranial static magnetic field stimulation (tSMS), have emerged as promising tools in neurorehabilitation. These methods modulate neural activity in targeted brain regions, enhancing neuroplasticity and functional recovery without the need for invasive procedures. TMS and tDCS are particularly effective in motor and cognitive rehabilitation, while tSMS offers benefits in reducing cortical excitability. The development of portable, user-friendly devices allows for at-home treatment, offering significant advantages such as increased convenience, cost-effectiveness, and personalized care. Home-based NIBS facilitates more frequent treatment sessions, enhancing compliance and motivation, while enabling remote monitoring by clinicians. As technology advances, the integration of NIBS into neurorehabilitation programs is likely to become more prevalent, providing patients with improved opportunities for recovery and quality of life.

Non-invasive brain stimulation (NIBS) approaches hold great potential for adjunct therapies in neurorehabilitation. However, several challenges exist that have limited more widespread adoption for clinical use. In this talk we briefly discuss the challenge of dose-control and precise delivery of stimulation, some of the conceptual challenges of different NIBS protocols, and potential solutions to optimize delivery of NIBS in clinical settings.

This paper outlines a treatment protocol for upper limb rehabilitation post-stroke, using static magnetic fields for 3 months alongside conventional rehabilitation and daily robotic therapy. To evaluate efficacy and placebo effects, three groups will be studied: two receiving magnetic treatment (one with sham treatment), and a control group without magnets. Patient progress will be monitored through clinical assessments, diagnostic imaging, and neurophysiological tests to provide a comprehensive view of improvement.

This study proposes a novel approach to assess muscle fiber type distributions using electrical stimulation and surface electromyography (sEMG). By applying electrical impulses to muscles with different fiber dominance, we examine the temporal and frequency characteristics of the resulting responses. Our results demonstrate that non-selective and synchronized muscle fiber recruitment induced by electrical stimulation can provide valuable insights into muscle composition, particularly in the context of fiber type dominance. The methodology presented here could be utilized for muscle monitoring during processes such as the decay of fast fibers in sarcopenia, enabling early detection of this condition and potentially enhancing interventions for muscle health in aging populations.

Functional Electrical Stimulation (FES) is a treatment for paralysis in central nervous system injuries that improves walking ability. This study evaluated the spatiotemporal impact of the H-GAIT neuroprosthesis in 5 subjects with chronic stroke, comparing unassisted and assisted gait recorded in VICON. Results showed a non-significant increase in step length, cadence, and gait speed with FES. The H-GAIT neuroprosthesis may positively affect gait, though limited use in a single session did not yield statistically significant changes.

Electrical stimulation has been shown to guide neural plasticity due to the interplay between afferent feedback and supraspinal input. Here, we assessed the short-term effects in spinal circuits of different peripheral electrical stimulation (PES) strategies, delivered at the common peroneal nerve below motor threshold for tibialis anterior. Specifically, we evaluated changes in reciprocal Ia inhibition (RI) before, right after, and 30 min after three different stimulation conditions during treadmill walking: 1) PES during stance phase; 2) PES during swing phase; 3) no stimulation. RI increased when PES was delivered during swing and decreased when PES was delivered during stance. On the other hand, RI remained unaltered when there was no stimulation. These results confirm phase-dependent modulation of RI when PES was delivered at different gait phases. These findings have implications for the use of PES in neurorehabilitation and imply that phase-dependent PES is necessary to guide spinal plasticity.

Peripheral electrical stimulation (PES) has been shown to boost activity-dependent neural plasticity by recruiting afferent fibers in conjunction with physiological activity. This study aimed to explore the acute changes in motor unit (MU) recruitment and derecruitment discharge rate (DR) following a 20-min PES session in healthy volunteers. High-density electromyography decomposition from flexor carpi radialis was used to identify MU DR patterns before and after PES. Recruitment and derecruitment DRs were increased when PES was delivered in-phase with muscle contraction. These results expand our knowledge on the effects of PES as a neural rehabilitation tool to treat motor disorders.

High-frequency alternating current stimulation can induce a rapid and reversible nerve block, specifically in motor fibers. Our objective is to investigate the effects of 10 kHz high-frequency alternating current on resting tremor versus sham stimulation in people with Parkinson’s disease (PwPD). We present preliminary results of a crossover, randomized trial involving PwPD with tremor. Active stimulation was applied at 10 kHz for 20 min over the upper limb and was compared with sham stimulation. A sample of n = 4 PwPD was included. The active intervention reduced the resting tremor in all participants, showing a reduction with respect to baseline during stimulation (5.7 × 10–3 G2/Hz; SD 7.7 × 10–3). However, the sham intervention reduced the resting tremor in only 2 participants, with a slight change during stimulation (0.6 × 10–3 G2/Hz; SD 7.1 × 10–3). This finding suggests modulation of resting tremor through a partial blockade of Aα nerve fibers. The use of KHFAC may reduce the impact of tremor on daily living activities, offering a safe and easily applicable intervention. Future studies should address the duration of KHFAC effects.

Parkinson’s Disease is considered the second most common age-related neurodegenerative disorder after Alzheimer’s Disease. One of its most disabling symptoms is Freezing of gait (FOG), which is a temporary, involuntary inability to step during initiation or turning while walking, typical in patients in a mid or advanced stage. The significant relationship between Parkinson’s FOG and gait characteristics motivates research on the influence of vibratory activation on proprioception during walking. The goal of this work is to present a vibratory system designed to provide gait-cycle-synchronised vibratory stimulation on lower limb muscles and reduce FOG occurrences. After technical validation, it is able to capture angular velocity data from both shanks and detect the main gait events to synchronize vibratory stimulation with the lower limbs muscles activation.

This study explores the phase-dependent effects of closed-loop transcutaneous spinal cord stimulation (tSCS) on the severity of tremors in patients with essential tremor (ET). Six ET patients (ages 53–80) were recruited to test a system integrating inertial measurement units and an Arduino microcontroller for phase-specific tSCS. The protocol included delivering tSCS at eight distinct phases of the tremor cycle and a no-stimulation condition. Results indicated a trend where tSCS applied in counterphase causes opposite effects (amplification vs reduction), though differences were not statistically significant (p > 0.05). These findings suggest that precise timing of tSCS may modulate tremor, highlighting the potential for personalized neuromodulation therapies. Further research is needed to confirm these results and optimize treatment strategies.

Burst nerve stimulation (BNS) increases corticospinal excitability producing increased motor output, and thus has been used for sensorimotor restoration in neurological disorders. It has been shown to provide a targeted effect to the muscles innervated by the stimulated nerve in the upper limb but a distributed effect in the lower limb muscles. However, all prior studies evaluated the impact of burst nerve stimulation on the muscles distal to the stimulation site in the upper limb. We aimed to examine the impact of BNS on the proximal muscles from the stimulation site in the upper limb. Motor-evoked potentials from seven healthy participants were recorded before and after BNS from the Flexor carpi radialis muscle. Our results indicate that burst nerve stimulation does not affect the corticospinal excitability in the target muscle, proximal from the site of nerve stimulation in the small sample tested.

In this case report we recorded Electroencephalograph (EEG) from an individual with complete spinal cord injury during spinal cord transcutaneous stimulation (scTS). The recording was done when stimulation was ON and OFF, and when the participant made an intention to move in the two stimulation conditions. Results show that stimulation correlates with an increase in EEG power in specific brain regions involved in movement processing and control, areas such as primary motor cortex, posterior cingulate cortex and supplementary motor area. The data suggests corticospinal engagement during spinal cord transcutaneous stimulation that could facilitate functional recovery and could inform rehabilitation programs for individuals with spinal cord injury.

Individualized spinal cord transcutaneous stimulation (scTS) parameters during overground gait training facilitate coordination in kinematics profiles in individuals with incomplete spinal cord injury (SCI). We extended our work to evaluate the immediate and within session motor response changes during gait, with and without stimulation. For gait trials with scTS, there was an immediate effect on the intralimb coefficient of variation for lower-extremity muscle groups compared to trials without scTS. Post training intervention, when trials of overground gait training combined with stimulation were compared to trials with stimulation at baseline, the results showed reduced interlimb coefficient of variation in muscles corresponding to spinal stimulation sites. These changes suggest that voluntary motor gains demonstrate the effectiveness of targeted scTS+training, and its sustenance post training.

The overall objective of this pilot study is to identify brain regions that characterize or manipulate cognitive effort in activities that involve both cognitive and motor efforts (dual-task), and to examine the neuromodulatory effect of different non-invasive neuromodulatory stimulation methods of the dorsolateral prefrontal cortex, transcranial alternating current stimulation (tACS) at 4 Hz and 30 Hz versus sham. The experiment utilized hand kinematic data collected from the LEAP motion controller along with EEG data to examine the effects of a custom-built car racing exergame on cognitive processes. Participants were instructed to memorize one or three objects and subsequently engaged in the game, which involved collecting coins, memorized objects, and avoiding traffic cones and novel objects. Preliminary results indicated higher desynchronization in sensorimotor regions during the condition with three objects, suggesting a greater neuromodulatory effect with increased cognitive effort.

In this paper, we propose to measure and characterize Electromechanical Delay (EMD) through a straightforward experiment that incorporates Functional Electrical Stimulation (FES) within a controlled motion environment. To this end, we have developed a data acquisition system specifically designed to synchronously detect stimulation events. The system was tested on both fatigued and non-fatigued quadriceps muscles. Our findings quantify the time between FES-induced and the torque response of the joint.

In this case report we recorded muscle activation from an individual with incomplete spinal cord injury after gait training alone and after 8 sessions of stimulation with gait training. After 60 sessions of gait training alone there was an increase in Co-Contraction of the agonist/antagonist motor pools that decreased after 8 sessions of transcutaneous spinal stimulation with gait training. The results show that stimulation with training increase inhibition for motor recovery.

This paper presents BCI-3D ReHaB, a rehabilitation system based on Brain-Computer Interface (BCI) technology that integrates Virtual Reality (VR) to aid in the recovery of motor impairments in stroke survivors. The system utilizes visual feedback and customizable 3D environments to enhance the rehabilitation process. The study focuses on developing an innovative User Interface (UI) that facilitates movement visualization and personalization. Additionally, the system incorporates electroencephalogram (EEG) acquisition to monitor real-time brain activity, ensuring the synchronization of EEG signals with physical movement markers. Pilot tests demonstrate the system’s effectiveness in aligning neural activity with motor actions and supporting motor imagery (MI). BCI-3D ReHaB represents a significant advancement in stroke rehabilitation, offering a personalized and immersive approach that aligns neural activity with physical movements, thereby enhancing the recovery process.

This study explores the feasibility of using wearable neurotechnologies to investigate neurofunctional patterns of interpersonal attunement during naturalistic coach-athlete interactions in sports. Eleven coach-athlete dyads participated in structured and unstructured feedback conditions, with EEG data collected using wearable EEG. Analysis focused on dissimilarity of alpha band responses, revealing significant differences in inter-brain attunement between conditions. Greater attunement was observed in the prefrontal and centro-temporal areas during structured interactions, while unstructured interactions showed attunement primarily in prefrontal regions. These results highlight the potential of wearable neurotechnologies for ecological hyperscanning, offering a novel approach to enhancing our understanding of communication dynamics in sports and improving coach-athlete relationships and performance.

Visual feedback is essential to maintain force steadiness during isometric contractions, but its effect depends on the specific feedback provided. Here, we used corticomuscular coherence (CMC) to investigate the effects of different visual gains (sensitivity of visual feedback at high, medium, and low levels) and the presence or absence of feedback on force control. Eight healthy adults were involved, recording the intramuscular electromyogram (iEMG) of the soleus muscle and electroencephalogram (EEG) of the primary motor cortex region. The results indicated that the alpha frequency band exhibited significant differences ( $$p<$$ p < 0.05) in CMC between high and low visual gain, and between the presence and absence of visual feedback. These findings highlight the sensitivity of the alpha band to visuomotor integration during isometric force motor control tasks.

This study explores the impact of salient stimuli on the beta frequency band at the muscle level during voluntary isometric contractions. Results showed that auditory stimuli consistently caused an early brief reduction in force immediately followed by longer-lasting increase and a subsequent increase in beta activity in muscle recordings. These findings suggest that auditory events can modulate motor cortical outputs, inducing beta activity that travels through the motor nervous system, offering a practical framework for studying cortical rhythms and potential strategies to modulate them.

Skill learning is a fundamental adaptive process, but the mechanisms remain poorly understood. Some learning paradigms, particularly in the memory domain, are closely associated with gamma activity that is amplitude modulated by the phase of underlying theta activity, but whether such nested activity patterns also underpin skill learning is unknown. Transcranial alternating current stimulation (tACS) over sensorimotor cortex is one way to modulate theta–gamma activity during motor skill acquisition, as an exemplar of a nonhippocampal-dependent task1. Indeed, this approach can significantly improve skill acquisition and provides a putative novel intervention for optimizing functional improvements in response to training or therapy. This talk will discuss related work including recent modelling approaches and highlight challenges and limitations.

Beta activity (13–30 Hz) has been consistently reported over the sensorimotor cortices before, during and after the completion of movements. Post-movement beta rebound is characterized by a rapid increase of power once an action is performed, but its functional role remains a matter of debate. Crucially, neural activity in the beta band has recently been acknowledged to occur as transient burst events rather than as sustained oscillations. Using a novel, adaptive burst detection algorithm and PCA-based waveform analysis, we show that bursts with different waveform shapes have differing rate dynamics during a joystick-based sensorimotor adaptation task. Focusing on post-movement bursts, we show that specific burst waveform shapes have distinct relationships with motor behavior. This study emphasizes that beta activity is best viewed transient bursts which diversity in shapes can reflect diversity in functions.

Beta band activity (13–30 Hz) is projected from the cortex to motor neurons and therefore to muscles. This transmission occurs primarily through the fastest corticospinal fibres. However, it is currently unknown whether this projection is uniform across all motor neurons innervating the same muscle. In this study, 12 subjects performed ankle dorsiflexion contractions while electroencephalography and high-density electromyography of the tibialis anterior muscle were recorded. Corticomuscular coherence was computed for increasing contraction force levels, which engaged motor units of progressively greater sizes. Beta band seemed to be uniformly transmitted to the muscle, as the average coherence did not significantly change across different force contractions. This suggests that beta oscillations project similarly to this muscle independently to the size of the motor neurons transmitting it.