Fig. 5

Neuro-cognitive assessment results A Grand average ERP response in channel Cz from all subjects combining gravity and myoprocessor control schemes for error vs. non-error trials in error blocks. Difference represents the difference between error and non-error trials. The response during the unassisted blocks is shown as a reference. Topographic visualizations of the difference between error and non-error trials in error blocks for gravity and myoprocessor control schemes respectively are shown below the ERP time-course plot. Regions outlined show areas of significant difference from Tukey HSD tests with \(p<0.05\). B Grand average motor response (biceps, triceps, and elbow angular velocity) from all subjects grouped by control scheme for error vs. non-error trials in error blocks. C Classification results from user experience neuro-cognitive assessment study. Trained rLDA models based on signals from EEG, EMG, and elbow angular velocity were used to classify control error trials during error blocks. The gravity and myoprocessor (Myo) control methods were compared to determine if accuracy was affected based on the control measure. Additionally, a combined accuracy for both control methods was determined per model. Error bars indicate mean ± SEM across subjects per model and control scheme. Representative confusion matrices for the combined control scheme models from subject 1 are depicted beneath the group results