Comparison between monopolar and bipolar cortical stimulation
Both monopolar and bipolar stimulations were performed through an ECoG electrode array under optical imaging. Figure 1 demonstrates typical reflectance changes evoked by the two stimulation methods. As expected, both stimulations evoked consistent reflectance trends over time, reaching their peaks at about 1.5 s after stimulation onset. Meanwhile, the time courses show larger peaking values evoked by bipolar stimulation than monopolar did, with the same stimulation parameters. The areas near electrode contacts (red dots near the solid arrows in Fig. 1A) show significant differences in reflectance changes. In addition ROI calculations from distal aretohowed no statistical differences between these two stimulation methods. The calculated results of ASC show a similar trend with that of MRC, with larger response areas evoked by bipolar stimulation than monopolar.
We then measured the effects of current intensity and frequency changes. As illustrated in Fig. 2, the two stimulation methods show similar trends in response to current intensity changes, with constant increases in MRCs resulting from increasing current intensity. Repeated-measures ANOVA reveals significant effects of both monopolar vs bipolar (p < 0.0001, F = 184.36), and current intensity (p < 0.0001, F = 194.56) in current modulation tests. Frequency modulation also shows significant effects (p < 0.001, F = 54.69). Post hoc testing with Bonferroni’s multiple comparisons shows that monopolar stimulations and bipolar stimulations differed significantly (p < 0.001). For current intensity, stimulation with 20 to 80 μA all show significant dereferences from each other, while stimulation with 80 μA and 100 μA shows no significant difference from each other. For frequency changes, stimulation at 20 Hz differed significantly from that at 10, 50, 100, 200 Hz, while others did not differ significantly from each other. To further discover whether the changing trends were consistent between the two stimulation methods, we calculated the changing ratios of monopolar to bipolar stimulations. The analytical results show that the ratios of intensity modulation remain within a stable range with no significant difference found, suggesting that the effect of current intensity modulation is not affected by the conditions of the electrode connections.
As to spatial distribution, we calculated the pixels of the areas having more than 3‰, 5‰, 7‰ reflectance changes (3‰, 5‰, 7‰ ARC for short, respectively) in each subject. For demonstrations of the differences between the two stimulation methods, we also calculated the relative changing ratios of monopolar to bipolar results. Figure 2C demonstrates the calculated results of changing area ratios. Compared to monopolar stimulations, bipolar stimulations resulted in larger changing areas in all the conditions we tested. Besides, monopolar stimulations with lower current intensities (20 and 40 μA) could hardly evoke strong reflectance changes (5‰ and 7‰ ARCs), and the area ratios also showed quick drops with larger reflectance changes, indicating that a threshold is required for activating brain activities to a certain extent. These results demonstrate a stronger and more focused reflectance effect evoked by bipolar stimulation. Notably, although the reflectance changes of monopolar stimulation seemed to be lower than that of bipolar stimulation, once reaching the threshold, the area ratio remained relatively stable with the increase of stimulation intensities.
Frequency modulation leads to different situations. As shown in Fig. 3, changes in frequency result in non-linear changes of reflectance, with peaking changes appearing around 100 Hz. However, the changing ratios of the two stimulations remain stable throughout the frequency changes, and no significant changes were found. As to changing areas, similar results were found with intensity changing conditions. Calculated results of ASC reveal similar trends to reflectance peaking values, whereas bipolar stimulation evoked a larger response of 5‰ and 7‰ ARC. These results again suggest a relatively more concentrated activation area with bipolar stimulation than monopolar stimulation.
Impacts of electrode spacing on cortical stimulation
Next, we measured the influences of electrode spacing. Two different spacing conditions were tested: two neighboring electrode contacts, and a doubled spacing pair of electrode contacts (spacing of 300 μm, and 600 μm, respectively). Reflectance changes show larger peaking values in response to the shorter spacing pairs of electrodes (300 μm), with significant differences between the two conditions (repeated measures ANOVA, p < 0.01, F = 44.92; Bonferroni’s multiple comparison, p < 0.01, Fig. 4B, C). Whereas significant changing areas show larger spread areas with the longer spacing condition (600 μm). Advanced analyses indicate that although larger electrode spacing enlarges spread areas, the area with 5‰ or higher reflectance changes were hardly found in large spacing conditions (Fig. 4D), suggesting that the activation of electrical stimulation is not only related to the current density but also related to the propagation distance. Comparatively, under short electrode spacing conditions, cortical responses reveal a more focused and balanced spread among the affected area (repeated measures ANOVA, p < 0.0001, F = 573.28; Bonferroni’s multiple comparison, p < 0.01). These results indicate that spatial specificity effects of electrical stimulation could be adjusted by the spacing of electrodes to some extent, the smaller the spacing, the stronger and more concentrated the response is.
Effects of multi-electrode stimulation
Researches have shown that properly patterned multi-electrode simultaneous stimulations may produce electrical fields with desirable spatial specificity. In the current study, quadrilateral configurations of the cathodic–pulse stimulation were applied in this part of the study. In 4 sources-1 return configuration, 4 electrode contacts delivered cathodic pulses simultaneously, and the central electrode was used as the return electrode. In 1 source-4 return arrangement, the centered electrode acted as a single source and the surrounding 4 electrodes were used as return electrodes. To eliminate deviation induced by total charge differences, current intensity applied in 4 sources-1 return arrangement was a quarter the intensity delivered to 1 source-4 returns condition. Imaging results show that reflectance responses are significantly stronger with 4 sources-1 return rather than 1 source-4 return arrangement (repeated measures ANOVA, p < 0.0001, F = 144.31; Bonferroni’s multiple comparison, p < 0.01, Fig. 5B, C. Response areas also show a significant difference between the two conditions, revealing a larger spread with 4 sources-1 return arrangement (repeated measures ANOVA, p < 0.0001, F = 412.74; Bonferroni’s multiple comparison, p < 0.01, Fig. 5D). These results indicate that multi-electrode simultaneous stimulations could modulate spatial spreading specifically.
Comparison between simultaneous and interleaved stimulations
Another stimulation condition recommended is interleaved stimulation. Interleaved stimulation with two pairs of electrodes could reduce charge accumulation and fatigue effect, and is considered safe for long-term stimulation. We compared responses to stimulations with one pair of electrodes, two pairs of electrodes simultaneously, and two pairs of electrodes interleaved. To ensure an equal total electric quantity delivery, one-pair stimulations were applied with 10 pulses, and two-pairs stimulations were applied with 5 pulses. Reflectance changes show significant differences among all three conditions, with the largest responses evoked by one single pair of electrodes, and the smallest responses responded to interleaved stimulations (Fig. 6B, C). Repeated measures ANOVA shows significant effects among the three conditions (p < 0.0001, F = 439.69); and Bonferroni’s multiple comparison reveals significant difference between one pair of electrode and two-paired interleaved stimulations (p < 0.01). As to response areas, the single-pair-electrode stimulation also showed the largest affected area, while the two-paired interleaved stimulation resulted in the smallest changing area. This may be due to the different stimulation durations (pulse numbers) of the different conditions. Repeated measures ANOVA shows significant effects among the three conditions (p < 0.0001, F = 655.80); and Bonferroni’s multiple comparison reveals more significant differences between one pair of electrode and two-paired simultaneous stimulations and between one pair of electrode and two-paired interleaved stimulations (p < 0.001), the two conditions of two-paired stimulations also show significant difference between each other (p < 0.01).These results show that, although the response intensity reduced, the interleaved stimulation remains relatively even spatial effects.