Water running in gait training and rehabilitation improves walking function in people with limited mobility, reduces impact on the lower limbs, provides horizontal upward force to support the body, and alleviates balance demands. Mechanical weight support mechanisms are similar to underwater walking, with underwater buoyancy simultaneously introducing greater resistance due to water viscosity, which is beneficial in gait rehabilitation. Traditional gait event detection relies on grounded counterforce measurements and video-based motion capture, although these methods lose their usefulness in the water running environment. Therefore, in order to continuously monitor gait in the field, researchers have explored the application of lightweight and wireless inertial measurement unit (IMU) sensors to enable gait event detection underwater.

The study provides a detailed analysis of the application of a machine learning based approach to foot IMU sensor data for gait event detection. In the study, subjects performed weight-assisted walking with mechanical support on a land-based treadmill and completed walking on an underwater treadmill. By using foot IMU sensors, gait event detection accuracies of up to 95-961 TP3T were achieved at different gait speeds and weight support levels, demonstrating the efficient performance of IMUs in combination with machine learning for assessing gait.