Control of trunk posture to improve gait rehabilitation

Recently featured in “Scientific Reports”, a rehabilitation robotic system that controls trunk posture in closed-loop improves locomotor performance during gait rehabilitation after spinal cord injury. To date, rehabilitation robotics has primarily focused on assistive devices that guide leg movements in order to maximize locomotor consistency and effort during training. Despite the importance of trunk posture on gait biomechanics, trunk assistance has comparatively received little attention. Typically, therapists immobilize the trunk or manually adjust pelvis movements. This study documents the substantial yet predictable impact of trunk posture on the modulation of bilateral leg kinematics and muscle activity. This knowledge is then leveraged to control trunk orientation and postural sway in real-time, which immediately improves stepping quality in rats with spinal cord injury. These results stress the importance of developing similar trunk assistance systems for humans.

Spinal cord injury (SCI) disrupts the communication between the brain and the sensorimotor feedback circuits in the spinal cord that coordinate leg movements. This disconnection interrupts the descending sources of excitation and modulation that are necessary to produce movements, leading to paralysis of the legs.

In order to promote neuroplasticity and functional recovery, gait rehabilitation has traditionally sought to provide repetitive task-specific leg movements, which in turn reinforce reproducible sensory feedback cues to spinal circuits and trains sensorimotor pathways. As a result, the design of training protocols, robotic interfaces and neuroprosthetic systems has been focused on guiding leg trajectories during rehabilitation. In these scenarios, the trunk is typically constrained in a bodyweight support system that provides vertically restricted forces.

However, natural locomotion involves precisely timed trunk movements in multiple directions, which directly determine leg biomechanics and loading, and consequently the modulation of leg sensory feedback circuits during locomotion. In this study, we show that trunk postural control induces robust, predictable effects on bilateral leg locomotor patterns through the modulation of proprioceptive afferent feedback circuits. We exploited these results to design closed-loop control strategies that regulate trunk posture and dynamical sway in real-time based on simple biomarkers. These control policies substantially improved locomotor performance in rats with severe SCI.

These results provide a conceptual and technical framework to design clinically-relevant assistive devices that tailor trunk posture in real-time in order to introduce precision medicine in gait rehabilitation after neurological disorders.

To read the publication, click here. 

To view the related video click here.