Related Experiment Video
Updated: Sep 27, 2026

A Rehabilitation Program of Exoskeleton-assisted Body Weight-Supported Treadmill Training with Non-immersive Virtual Reality for Stroke Patients
Published on: May 16, 2025
A Multimodal Therapist-Supervised Robotic Gait-Training Platform with Phase-Synchronized Infrared Thermal
Rocco Salvatore Calabrò1, Aurelio Crespantini2, Andrea Calderone1
1IRCCS Centro Neurolesi Bonino Pulejo, 98124 Messina, Italy.
Abstract:
Robotic gait technologies can deliver intensive, repetitive, task-oriented stepping, but multimodal platforms require careful integration before clinical testing. This concept-level technical proof-of-concept describes Li-Walk®, an investigational fixed robotic gait-training platform combining a treadmill-synchronized lower-limb exoskeleton, dynamic body-weight support, phase-synchronized infrared-A (IR-A) thermal biofeedback, directional acoustic feedback, a semi-immersive display, camera-derived body representation, and artificial intelligence (AI)-supported therapist guidance. The architecture is defined through subsystem interfaces, implementation status, operational states, explicit risk controls, and a staged validation roadmap. No human participants, patient-level data, bench datasets, or statistical analyses were involved; subsystem factory checks did not characterize integrated performance. The installed exoskeleton has four actuated sagittal axes at the hips and knees, with in-series load cells supplying interaction-force inputs for contingent ipsilateral IR-A cueing under reduced guidance. The integrated AI module combines rule constraints with a case-based design, but has no trained statistical model or clinical cohort database; the therapist retains parameter-setting authority. Live video and avatar modes are implemented, whereas independent thermal monitoring and structured recommendation audit trails remain planned. Trigger thresholds, sampling rates, end-to-end latency, exposure limits, and fault responses remain undocumented or unverified. The contribution is a conceptual integration framework that distinguishes implemented functions, declared specifications, and unverified requirements. Independent mechanical, thermal, software, and human-factors validation, followed by appropriately authorized human studies, is required to establish safety, usability, technical reproducibility, and any incremental rehabilitation benefit. The optional thermal branch requires particular safeguards for impaired thermal sensation and cannot be assumed to add clinical value beyond robotic gait training alone.
