Fig. 10
From: User training for machine learning controlled upper limb prostheses: a serious game approach
SHAP scores for the spherical subset of the SHAP per participant. Higher scores reflect better performance. Participants in the game group are in solid colour (P3, P6), while participants in the conventional group are shaded (P2, P5). Baseline was conducted before training using the participants’ own DC controlled prosthesis. Pre- and post-tests were conducted using the Michelangelo hand controlled with ML. In the pre-test all participants used one DoF and in the post-test participants used two DoFs