Figure 2:

iBCI robustness to spike error. Data recorded during centre-out-and-back cursor trials of the three monkeys (J, L, and R) and the human participant (T5). (a) SER (spike errors/sec) simulation process. First, binary threshold neural signal bits (’1’ - spike , ’0’ - otherwise) were flipped at different rates (10⁻⁶ to 10^−0.5). Second, cursor velocity was predicted from the noisy signal. (b) Decoder performance (velocity coefficient of determination, R²) as a function of added SER (ranges from 0 to 1). Values are normalized to performance when decoding an undistorted signal (i.e., $\frac{R_{SER}}{R_{original}}$). Vertical bars along the lines represent the standard error of 10-fold cross validation across 10 days (total of 100 R² estimates). Horizontal bars in the right top corner indicate significant change in performance compared to the undistorted signal (two-sided Wilcoxon rank-sum test, p<0.05). Vertical dashed line indicates the SER in which performance starts to degrade. Bar colors correspond to the user.