Billot, M. | Search bioRxiv

bioRxiv2026

The influence of virtual visual stimulus amplitude to induce standing postural responses

Upright postural control during movement relies on multisensory integration. Yet, the frequency-specific contribution of vision remains poorly characterized in virtual reality (VR). This study investigated how multi-sine visual stimulation amplitude delivered in VR influences standing postural responses. Fifteen healthy adults stood on a force plate wearing a VR headset. Visuo-postural coupling was assessed through coherence and gain analyses between a multi-sine signal (10 sinusoids, 0.12 to 1 Hz) oscillating a virtual environment in one of four amplitudes (0.5, 1, 2, 4 degrees peak-to-peak) and the anteroposterior whole-body angle. All amplitudes elicited measurable postural responses. Increasing amplitude significantly increased postural oscillation and tended to increase coherence, while gain significantly decreased. These results are consistent with a nonlinear control system. The 2 degrees amplitude elicited the largest gain with significant coherence across all stimulated frequencies, suggesting it is suitable for studying visual contributions to postural control during movement execution.

neuroscience↗

bioRxiv2025

MyoBack: A Musculoskeletal Model of the Human Back with Integrated Exoskeleton

Given the challenges of real-life experimentation, musculoskeletal simulation models could become essential in biomedical research. This is especially critical for the human back, a key structure involved in daily movements, where modeling and simulation could streamline design and support the development of treatments and robotic rehabilitation techniques, such as exoskeletons. However, musculoskeletal simulation engines are computationally demanding and lack contact dynamics, restricting current models use in studying prolonged behaviors or optimizing system design while maintaining physiological accuracy. To overcome this limitation, this work proposes MyoBack, a human back model part of the MyoSuite framework relying on the physics engine MuJoCo. This model is derived from a physiologically accurate model built in the state-of-the-art musculoskeletal simulation software OpenSim and replicates the latters kinematic properties accurately, with some discrepancies regarding muscle dynamics stemming from engine differences. The MyoBack model was also validated empirically by integrating a passive back exoskeleton in simulation and comparing forces exerted on the back with values from experimental trials. Over different tasks, the model reproduced measured force progressions well, resulting in RMSE = 11% for a stoop and RMSE = 16% for a squat motion pattern relative to peak forces. The MyoBack model can be accessed here: https://github.com/rohwalia/MyoBack

bioengineering↗