Upper limb impairment and spastic hypertonia following a stroke often necessitates constant wrist stretching for symptom relief and functional recovery. Most existing rehabilitative devices, however, can be difficult for stroke survivors to don and operate with one functional hand, which creates a barrier to at-home, self-administered therapy. To address this, we present SUREHand, a novel, passive wrist exoskeleton specifically designed for hemiparetic stroke survivors to independently don and use with one hand. The device features a pneumatic support system with an inflatable air bladder to ensure a secure, comfortable fit and a body-powered ratchet mechanism that allows users (MAS 1+) to manually induce and maintain wrist extension. We recruited three stroke survivors (N=3) to evaluate the device’s mechanical performance and device usability. All participants operated the device independently in under 20 minutes of training, with average donning times of 72 ± 38 seconds and wrist extension with the device achieving their maximum passive range of motion. High scores on the System Usability Scale and comfort survey suggest that the SUREHand is an effective, low-cost, and accessible solution for wrist rehabilitation and stretching.
In teleoperation, the lack of tactile feedback limits an operator’s ability to accurately assess remote surface conditions, affecting manipulation precision. This paper presents an experimental study investigating how tactile feedback and different control modes affect human operators’ ability to identify remote environments during teleoperation. Two interaction strategies were compared: teleoperation, in which users controlled robot motion while receiving contact-related vibrotactile feedback, and auto-operation, in which the robot followed predefined trajectories and users received tactile feedback only. Eighteen participants performed remote identification tasks. The results show that participants could interpret simplified vibrotactile cues in both strategies, achieving comparable identification accuracy during active control and passive sensation. This result indicates that decoupling motor control from haptic sensing may not impair perceptual performance, offering flexibility in teleoperation interface design where operators can focus on perception while robots execute exploratory movements autonomously.
Personalized wearable devices (orthoses, exoskeletons, prosthetic sockets) require accurate geometric representations of patient anatomy, yet current workflows rely on manual CAD modeling. We present an automated pipeline that learns sparse parametric representations directly from 3D scans. Our method combines an optimized anchor-based patch decomposition with PointNet-based control point regression to transform dense point clouds (10k-400k points) into compact B´ezier parameters (16 control points per patch). Evaluation on clinical scans achieves mean reconstruction errors below 12mm while maintaining CAD compatibility. The resulting low-dimensional representations enable rapid design iteration, seamless fabrication workflows, and compatibility with physics simulators for robotic applications. We further compare against a closed-form least-squares baseline on the same patches and demonstrate that the learned estimator provides a bounded worst-case error (Hausdorff one to three orders of magnitude lower).
Functional Electrical Stimulation (FES) can support hand movement after neurological injury, but multi-channel systems remain difficult to operate because low-bandwidth user interfaces must control high-dimensional stimulation patterns. We propose a Gaze-Driven Perception-Action Control Framework that transfers part of the high-level selection process to an autonomous pipeline. Temporally stable, task-conditioned gaze behavior is used to trigger grasp selection, while object semantics and hand kinematics determine the channel-specific stimulation strategy. A pilot feasibility study with three healthy volunteers was conducted in controlled tabletop tasks. Relative to a manual UI baseline, the proposed framework reduced the overall mean end-to-end task completion time by approximately 17%, with object-wise reductions of up to 28.4%, and maintained relatively stable interaction time as the number of available gesture options increased. These results provide preliminary evidence that gaze-driven control can improve interaction efficiency and reduce manual selection demands in multi-channel FES. The study does not establish clinical effectiveness; validation in neurological patients and more realistic environments is required.
This paper presents EchoGym, a real-time, bidirectionally synchronized digital twin (DT) framework for robotic echocardiography. EchoGym is designed to support data-driven algorithm development, sim-to-real evaluation, and operator training when physical experimentation is limited or impractical. The framework integrates a physical robotic system, a physics-based simulation in NVIDIA Isaac Sim, and a generative AI-based ultrasound image-synthesis module through consistent robot operating system 2 (ROS 2) interfaces. Robot joint-velocity commands and ultrasound imaging parameters are processed to update the virtual system and generate cardiac ultrasound images conditioned on the corresponding probe pose and imaging settings. Image synthesis follows a sample-efficient two-stage approach. First, convolutional neural network decoders generate low-resolution ultrasound images at discrete imaging-parameter settings, and trilinear interpolation produces images at intermediate settings. A super-resolution generative adversarial network (SRGAN) then enhances the generated images. The key capabilities of EchoGym are evaluated through hypothesis-driven experiments. The findings demonstrate real-time, high-quality ultrasound-image generation, accurate low-latency physical–virtual synchronization, successful sim-to-real transfer of UIBVS, and effective SAC-based probe-motion guidance during joystick-controlled robotic scanning, establishing EchoGym as a unified DT platform for data-driven robotic echocardiography.
