Effects of tDCS dose and electrode montage on regional cerebral blood flow and motor behavior

Abstract

We used three dose levels (Sham, 2 mA, and 4 mA) and two different electrode montages (unihemispheric and bihemispheric) to examine DOSE and MONTAGE effects on regional cerebral blood flow (rCBF) as a surrogate marker of neural activity, and on a finger sequence task, as a surrogate behavioral measure drawing on brain regions targeted by transcranial direct current stimulation (tDCS). We placed the anodal electrode over the right motor region (C4) while the cathodal or return electrode was placed either over a left supraorbital region (unihemispheric montage) or over the left motor region (C3 in the bihemispheric montage). Performance changes in the finger sequence task for both hands (left hand: p = 0.0026, and right hand: p = 0.0002) showed a linear tDCS dose response but no montage effect. rCBF in the right hemispheric perirolandic area increased with dose under the anodal electrode (p = 0.027). In contrast, in the perirolandic ROI in the left hemisphere, rCBF showed a trend to increase with dose (p = 0.053) and a significant effect of montage (p = 0.00004). The bihemispheric montage showed additional rCBF increases in frontomesial regions in the 4mA condition but not in the 2 mA condition. Furthermore, we found strong correlations between simulated current density in the left and right perirolandic region and improvements in the finger sequence task performance (FSP) for the contralateral hand. Our data support not only a strong direct tDCS dose effect for rCBF and FSP as surrogate measures of targeted brain regions but also indirect effects on rCBF in functionally connected regions (e.g., frontomesial regions), particularly in the higher dose condition and on FSP of the ipsilateral hand (to the anodal electrode). At a higher dose and irrespective of polarity, a wider network of sensorimotor regions is positively affected by tDCS.

Keywords: Arterial spin labeling; Bihemispheric electrical stimulation; Motor learning; Neural excitability; Sensorimotor network; rCBF change; tDCS.

Fig. 1.

Schematics of experimental setup. (a)Head models show the two electrode montages used in this study and stimulator shows different dose levels. (b) Experimental setup for concurrent tDCS-fMRI using the RF filter panel to transmit the current into the MR room. (c) Finger-Key Mappings for left and right hand. Top row shows original keyboard keys with number pads pasted on the those keys displayed in the second row. (d) Behavioral task timing diagram: Finger sequence task performed before and after tDCS, the sessions consisted of 5 epochs (E=Epoch of 30s) for each hand before and after the tDCS stimulation; the hand that performed the finger sequence task first was counterbalanced across participants. (e) Concurrent tDCS-fMRI timing diagram: 24 min of continuous rs-FMRI was recorded during which 10 min of tDCS was applied from minute 6 to minute 16. The rs-fMRI timing diagram shown in 1e was the same for the ASL and the BOLD experiment which were done in one session, but the order of the ASL or BOLD experiment was counterbalanced as to which one occurred as first or second experiment.

Fig. 2.

Percentage change in Finger Sequence Performance (FSP) comparing before to after stimulation (MEAN +/− SEM). (a)Left hand FSP change and (b)Right hand FSP change.

Fig. 3.

ASL T contrast values in Left (3a) and Right (3b) perirolandic ROI (MEAN + /- SEM) comparing ON versus OFF periods of tDCS stimulation.

Fig. 4.

tDCS tolerability analysis in a) behavioral study and b) concurrent tDCS-MRI study.

Fig. 5.

SPM Activation maps for all the active stimulation conditions with ON>OFF stimulation function and for dose-montage effects; (a, c, and e) Unihemispheric montage: SPM maps for 2mA, 4mA, and 4mA vs. 2mA dose. (b, d, and f) Bihemispheric montage: SPM maps for 2mA, 4mA, and 4mA vs. 2mA dose. g) Effect of Dose with function 4mA>2mA, and h) Effect of Montage with function Bihemispheric montage>Unihemispheric montage. Uncorrected threshold of p = p<0.05 was used to identify positive and negative changes in ASL represented by Yellow and Blue voxels, respectively. Red and Green voxels represent positive and negative changes in ASL values for p-values between 0.05 and 0.1).

Fig. 6.

Correlations between observed output variables. a) left and right ROI ASL T-contrast values against FSP plotted across all dose-montage combinations. (b) Tolerability scores against ASL T-Contrast and FSP values. (c) FSP values from left and right hand against simulated current intensity from left and right perirolandic ROI. (d) left and right ROI ASL T-Contrast against simulated current intensities. *- multiple comparison corrected significant correlation.