Victor R. Barradas

This paper introduces a hierarchical Bayesian framework to estimate subject-specific muscle parameters using EMG and torque data, achieving high accuracy and robustness in musculoskeletal modeling.

The paper shows that a 40% shoulder force reserve predicts compensatory trunk movements during reaching with shoulder strength deficits, outperforming traditional control models.

This study develops a neuro-musculoskeletal model to show how compensatory motor areas are recruited post-stroke to maximize residual motor performance through homeostatic and reinforcement learning processes.

The paper analyzes the speed of learning inverse models in motor tasks using theoretical and experimental approaches, finding that task geometry and target distribution influence learning speed.

This study shows that residual muscle activity not explained by synergy extraction significantly affects motor task performance, challenging the assumption that it is noise.