Fig. 2
From: Bypassing stroke-damaged neural pathways via a neural interface induces targeted cortical adaptation

Task performance improves with practice. a Task performance during the initial and last 2 min of a session and during the catch trials (N = 3 monkeys, n = 108 ACMC sessions, consisting of 72 torque-matching and 36 position-matching task sessions; Monkey TA, n = 60 sessions; Monkey M, n = 37 sessions; Monkey TE, n = 11 sessions). Bars indicate mean values and circles indicate individual data points. b Time course of averaged task performance across 108 sessions in the three monkeys; Black dots show mean values; thin lines represent individual sessions. Black and red asterisks represent significant differences (P < 0.05 by one-way ANOVA with Bonferroni’s correction for post hoc multiple comparisons) compared to the initial and last 2 min of the session, respectively. c Electrode locations. In Monkeys TE and M, the electrocorticography (ECoG) electrode array was placed on the surface of the frontal eye field (FEF; grey), premotor cortex (PM; yellow), primary motor cortex (M1; red), and primary somatosensory cortex (S1; blue). In Monkey TA, 16 microelectrodes were implanted within the hand area of M1 (red dots). The electrodes that were used as the input signal for the ACMC are encircled in black. Ars: arcuate sulcus, CS: central sulcus. d Time course of task performance using the ACMC with input electrodes in the FEF (grey line, n = 1 session in Monkey M), PM (yellow line, n = 18 sessions in Monkey TE and M), M1 with ECoG array (red solid line, n = 24 sessions in Monkey TE and M), M1 with microelectrode array (dashed red line, n = 60 sessions in Monkey TA), and S1 (blue line, n = 5 sessions in Monkey M). Task performance of M1 electrodes with ECoG array was significantly higher than that with the other areas, and task performance with FEF electrode was significantly lower than that with the other areas (P < 0.05 by two-way ANOVA with Bonferroni’s correction for Post hoc multiple comparisons). Detailed statistical results for (b), (d) are shown in the source data. Data in (d) are shown as means and standard errors