Yue Wen

This study evaluates how kernel dimensionality in CNNs affects neural drive estimation from HD-sEMG, finding that 2D CNNs offer the best balance of accuracy and efficiency.

This study evaluates the consistency of 7 open-source motor unit decomposition tools using HD-sEMG data, finding significant variability in results and requiring manual editing for reliable analysis.

This paper introduces a deep CNN for estimating neural drive from HD-EMG signals, showing it generalizes across muscles, sessions, and participants without retraining.