Abstract
We establish the validity of error-based and usedependent learning models to describe the response of healthy individuals to robot-aided training of propulsion via a robotic exoskeleton, and propose a modified UDL model that accounts for both use-dependent and error-based learning during and after training.
We formulated five different state-space models to describe the stride-by-stride evolution of metrics of propulsion mechanics (hip extension – HE - and propulsive impulse - PI) during and immediately after exposure to robot-assisted training, applied by a unilateral hip/knee robotic exoskeleton for 200 consecutive strides. The five state-space models included a single-state model, a two-state model, a two-state fast and slow model, a UDL model, and a modified UDL model requiring 4, 9, 5, 3, and 4 parameters, respectively. The coefficient of determination (R2) and Akaike information criterion (AIC) values are calculated to quantify the goodness of fit of each model during and after exposure to robotic training. Model fit was conducted both at the group level and at the individual participant level, for both outcome measures.
At the group level, the modified UDL model shows best goodness-of-fit compared to other motor adaptation models in AIC values in 15/16 conditions. For participant-specific level, both the modified UDL and the complete two-state model have significantly better goodness-of-fit compared to the other models. In summary, the modified UDL model is a simple 4- parameter model that achieves similar goodness-of-fit compared to a complete two-state model requiring 9 parameters. As such, the modified UDL model is a promising model to describe the stride-by-stride response of propulsion mechanics in robotaided gait training.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
* This work is supported in part by NSF-CMMI-1934650 and in part by NIH-R01HD111071.
fabs{at}udel.edu
secretgh{at}udel.edu