A Sensorized Insole to Estimate Ground Reaction Forces and Center of Pressure During Gait

Traditional methods for gait analysis require stationary equipment, leading to limitations in mobility and real-world applicability. Wearable devices, particularly sensorized insoles, offer a promising solution for gait analysis during sports and dynamic activities. This study presents the development and evaluation of a custom-made sensorized insole for accurate estimation of Ground Reaction Force (GRF) and Center of Pressure (CoP) during human gait. The insole integrates pressure sensors and an accelerometer, coupled with Long Short-Term Memory (LSTM) models, to capture temporal dynamics in gait patterns. Data from eleven healthy adult volunteers were collected, pre-processed, and used for model training and validation. Results demonstrate the effectiveness of the sensorized insole, with improvements in prediction accuracy when combining pressure and acceleration data. The addition of accelerometer data to pressure data led to a reduction in the Normalized Root Mean Square Error (NRMSE) for the medio-lateral component of GRF from 18.4% to 16.7%. While challenges remain, particularly in modeling medio-lateral components of GRF, the study provides insights into potential future directions for optimizing sensorized insoles and improving model performance.