How treadmill training rewires the walking network after spinal cord injury

What is it about?

Spinal cord injuries severely disrupt walking, but traditional scoring methods often miss subtle improvements in how the body coordinates movement. In this study, we used artificial intelligence to precisely track the joint movements of rats recovering from a spinal cord injury, combining this data with nerve and muscle measurements. Instead of looking at isolated metrics, we used a "network" approach to see how different body systems communicate with each other. We found that treadmill rehabilitation didn't just improve individual walking steps; it actually helped reorganize the connections between nerve reflexes and walking patterns, making the body's movement network look much more like that of a healthy, uninjured animal. This demonstrates that evaluating the "big picture" of body coordination is a much more powerful way to measure rehabilitation success.

Featured Image

Why is it important?

Traditional assessments of spinal cord injury recovery rely on coarse behavioral scores that miss subtle, system-level changes. Our work uniquely shifts the evaluation paradigm from single-metric scoring to a multi-domain network physiology approach. By combining AI-driven pose estimation with electrophysiology, we demonstrate that rehabilitation fundamentally reorganizes the cross-domain statistical associations between spinal reflexes and gait kinematics. Furthermore, we reveal that training specifically restores the "pattern formation" layer of spinal circuits, normalizing step regularity even when maximum speed remains constrained. This timely, data-driven approach provides a highly sensitive, network-level biomarker for rehabilitation effectiveness, offering a more accurate way to evaluate true functional recovery and bridge the translational gap.

Perspectives