Body-machine interfaces (BoMIs)-systems that control assistive devices (e.g., a robot) with a person's movements-offer a robust and non-invasive alternative to brain-machine interfaces for individuals with neurological injuries. However, commercially-available assistive devices offer more degrees of freedom (DOFs) than can be efficiently controlled with a user's residual motor function. Therefore, BoMIs often rely on nonintuitive mappings between body and device movements. Learning these mappings requires considerable practice time in a lab/clinic, which can be challenging. Virtual environments can potentially address this challenge, but there are limited options for high-DOF assistive devices, and it is unclear if learning with a virtual device is similar to learning with its physical counterpart. In this project, we are creating a novel virtual robotic platform for controlling high degrees of freedom assistive devices with body-machine interfaces. This platform could serve as a low-cost tool for learning to control assistive devices and for conducting motor learning experiments.

Individuals with stroke often experience significant impairments in their upper and lower extremities, affecting their daily activities and quality of life. Functional electrical stimulation (FES) can be used to treat these impairments and improve motor function after stroke. Typically FES are triggered using electromyographic (EMG) signals from the targeted muscles; however, continuous FES control using EMG signals is not feasible due to stimulus artifacts. Kinematic sensors, like bend sensors controlled by the contralateral hand, have shown effectiveness in hand function recovery after stroke. Despite this, there are no low-cost, open-source, modular interfaces versatile enough for various activities.

This project introduces an IMU-based controller for FES, which will soon be available as an open-source application. The NeuRRo-FES system is designed for motion-based FES of upper or lower limb muscles, using low-cost wireless IMUs placed on adjacent limb segments to capture orientation and calculate joint angles in real time. NeuRRo-FES can monitor up to three joints per limb, detecting movements and triggering or stopping FES based on customizable thresholds. The system includes a graphical user interface (GUI) for real-time adjustments of threshold, frequency, and intensity, allowing adaptability to different impairment levels. It also features velocity thresholds to reduce noise and cross-talk, ensuring near-perfect accuracy in muscle stimulation timing.

Neurological injuries can lead to significantly life-altering impairments of neuromuscular pathways. Paresis is a very common side effect of stroke, where the weakening of the muscles on one side of the body can lead to challenges in performing day-to-day tasks. Some current methods of robotic rehabilitation are expensive, electrically powered, and may not be readily accessible for frequent use. In this project, we are developing a lightweight, wearable, passive robot called the 'Hand eMbot'. This device uses a Bowden cable pulley mechanism to mirror the movements of one hand onto the other. By replicating the motion of the unaffected hand, the device assists in re-establishing neuromuscular pathways on the impaired side. We anticipate that this coordinated bilateral movement could help individuals regain strength, range of motion, and fine motor control skills. The device is also designed to work in combination with the NeuRRo FES system to further strengthen the neuromuscular response.

Our lab has developed several fully passive and semi-passive devices for gait rehabilitation. These devices use eddy current brakes to provide velocity-dependent resistance during walking. The short-term and long-term effects of these devices are being tested in able-bodied as well as in individuals with neurological or orthopedic injuries.

Gait impairments are a prominent source of disability after neurological or orthopedic injuries. Applying functional resistance training during walking is an emerging method for treating individuals with gait impairments. This training is administered by having a patient perform a task-specific training (in this case, walking) while a load is applied to resist the movement. This works simultaneously to improve muscle strength and coordination, which are often underlying sources of gait impairment. While there are several robotic solutions for this type of training, these devices are often too expensive for use in small clinics or in the home. Hence, we developed a unique, low-cost, passive exoskeleton to provide different types of elastic resistances (i.e., resisting flexion, extension, or bidirectionally) to the leg muscles during walking. This device uses a system of counteracting compressional springs, pulleys, and clutches to provide these different types of elastic resistance. We have shown that this device could specifically target knee flexors, extensors, or both, and increase eccentric loading at the joint. Additionally, these resistance types elicited different kinematic aftereffects that could be used to target user-specific spatiotemporal gait deficits. Hence, the device provides a potential low-cost option for addressing user-specific muscle weaknesses and gait deficits during functional resistance training.

In this project, we seek to determine how functional resistance training using different resistance types acutely promotes biomechanical and neurophysiological adaptations during gait in able-bodied and individuals with stroke or ACL-reconstruction.

In this project, we use eddy current brakes and exploit kinematic redundancies to provide direct resistive forces during planar reaching. Because there are no active actuators (i.e., motors) and the device can be mounted easily on a simple table, the device is expected to be more affordable and safe for in-home use.

In this project, we are creating an open source, virtual reality system for upper- and lower-extremity rehabilitation. The virtual reality system (NeuRRoVR) has several in-built games and interfaces to provide several existing and novel therapies (e.g., mirror therapy,  bimanual therapy, hand therapy, balance and coordination training) and also perform neuromuscular evaluations (e.g., Box and Block Test, joint excursions).

In this project, we use a mechanical perturberator to provide controlled perturbations to the ankle during standing and walking and use Least-squares System Identification techniques to quantify the mechanical impedance of the ankle. We are evaluating the effects of muscle activation, ageing, stroke, and Botulinum Neurotoxin on ankle mechanical impedance during standing and stance phase of walking.

In this NIH-funded, Phase-I STTR, NeuRRo Lab and investigators at the University of Michigan has partnered with Elite Athlete Products and Prof. Washabaugh at Wayne State University to refine, develop, and test a low-cost, gait and balance rehabilitation system called NewGait based on end-user feedback, biomechanical simulations, and rigorous scientific experiments. More information about this project can be found at https://reporter.nih.gov/search/WPztW9JB30GMav8RkHkPjQ/project-details/10611686