Skip to main content
. 2017 Dec 7;7:17149. doi: 10.1038/s41598-017-17464-1

Figure 3.

Figure 3

Performance of the myokinetic experimental prototype: Emodel, Ecross-talk, Esensor. Each MM was moved along the anatomically appropriate ROM of the relative muscle (a: MM0 – APL; b: MM1 – FPL; c: MM2 – FDS1; d: MM3 – FDS2), while the remaining MMs were kept at their initial/rest position. (A) Estimated versus actual displacements of the MMs: while the moving MM is appropriately tracked (R2 = 0.99) the computed displacement of non-moving MMs represent the cross-talk errors, Ecross-talk (dot markers). For each moving MM,the difference between the actual and the estimated displacement (Emodel) is shown with cross markers. (B) Emodel and Ecross-talk averaged over the entire ROM. Emodel and Ecross-talk correspond to the diagonal and non-diagonal elements of the confusion matrix, respectively. (C) Technological limitations affecting the myokinetic interface (Esensor). For both panels (B) and (C): each row contains the errors related to a single MM; each column contains the errors generated by the movement of a specific MM.