Fig. 6: Brain-controlled LHA-DBS enhances locomotor recovery after sever SCI.

a The schematic diagram of LHA DBS triggered by BCI. A 4 * 8 array was used to record neural signals from M1. The multi-unit activity (MUA) was processed online by an unsupervised algorithm. DBS application and intensity depended on the neural activities. b Representative leg kinematics, neural activities, and hindlimb EMG data of healthy mice (left) and staggered mice (right). Scale bar of representative neuronal active channel, 0.2 mV. Scale bar of EMG data, 0.4 mV, 0.5 s. c Trajectory of intact mice in open-field chambers without DBS (left), with irregular DBS (middle), and with brain-controlled DBS (right). d Trajectory of mice after SCI (stagger injury) in open-field chambers without DBS (left), with irregular DBS (middle), and with brain-controlled DBS (right). e Quantification of the proportion of center time (left), total distance (middle), and entries to center area (right) in intact mice, comparing no DBS to irregular or brain-controlled DBS (n = 6). One-way ANOVA/Tukey’s multiple comparison was applied. ns, not significant; *p < 0.05; **p < 0.01. Error bars, SEM. f Quantification of the proportion of center time (left), total distance (middle), and entries to center area (right) in post-SCI (stagger injury) mice, comparing no DBS to irregular or brain-controlled DBS (n = 6). One-way ANOVA/Tukey’s multiple comparison was applied. ns not significant; ****p < 0.0001. Error bars, SEM. g Timeline summarizing the experiments about the effect of LHA brain-controlled DBS in the recovery process of mice following stagger SCI. h Chronophotography of mice illustrates the recovery process following a stagger injury. i BMS test performance in control, BCI-DBS training without stimulation, and BCI-DBS training with stimulation groups. n =  5 per group, two-way ANOVA-RM with Bonferroni post hoc correction. **p < 0.01, ****p < 0.0001. Error bars, SEM. j Quantification of maximal crest height (left) and stride length (right) of mice at 70 DPI without training, trained without DBS, or trained with DBS (n = 5). One-way ANOVA/Tukey’s multiple comparison was applied. **p < 0.01; ***p < 0.001; ***p < 0.0001. Error bars, SEM. k Representative leg kinematics and hindlimb EMG data from mice without training (left), trained without stimulation (middle), and trained with stimulation (right) after stagger injury. Source data are provided as a Source data file.