Effect of left DLPFC electric field magnitude on tDCS-induced resting brain connectivity changes
Poster No:
102
Submission Type:
Abstract Submission
Authors:
Eunkyung Kim1, Seo Jung Yun1, Byung-Mo Oh1, Han Gil Seo1
Institutions:
1Seoul National University Hospital, Seoul, Korea, Republic of
First Author:
Co-Author(s):
Introduction:
Although transcranial direct current stimulation (tDCS) is known to be effective for modulating cortical activity, there is considerable variability in response and lack of understanding how resting state functional connectivity (rsFC) changes. The aim of this study was to investigate how variations in electrical field magnitude (E-field) applied to the target area of tDCS, which was the left dorsolateral prefrontal cortex (L-DLPFC), affect L-DLPFC-based rsFC changes in healthy adults.
Methods:
A double-blind, sham-controlled, counterbalanced cross-over design was applied on 21 healthy individuals (37.6±8.6y). Participants received either constant 2 mA anodal or sham tDCS targeting the L-DLPFC during 10 min (wash-out period; at least 3 d, Mindd Stim, Ybrain, Inc, South Korea). The resting-state fMRI (3-T) was acquired before and after stimulation. The L-DLPFC was localized in each individual (Mylius et al., 2013) and E-field (V/m) was estimated. A spherical region of interest (ROI) surrounding individual peak of the L-DLPFC (radius 10 mm) was generated to construct seed-based rsFC. The subject-level regression coefficients (COPEs), representing rsFC strength, fed into a higher-level group analysis.
To examine alterations in overall L-DLPFC rsFC after active and sham tDCS, average beta-estimates for both positive and negative L-DLPFC rsFC were extracted from the subject-level COPEs images within each condition. These estimates were obtained after masking the COPEs images by the thresholded connectivity map derived from all pre-stimulation conditions. Repeated measure of analysis of variance (RM-ANOVA) and paired-sample t-tests was conducted to test the significance of difference using MATLAB. To assess locally distributed L-DLPFC rsFC changes, mixed-effect analysis was conducted using FSL FLAME1, treating participants as random effects. Pairwise post-hoc comparison was also performed.
The relationship between the E-field delivered to the L-DLPFC and changes in L-DLPFC rsFC was investigated. The positive connectivity map of active tDCST0 (cluster-extent based thresholding |z| > 3.1, FWE p < 0.05) revealed 9 clusters, while the negative connectivity map revealed 14 clusters. Average beta-estimates of 23 clusters were extracted from the subject-level COPEs images of both active tDCST0 and tDCST1.The difference in beta-estimates between active tDCST0 and tDCST1 was correlated with the E-field while controlling for age, sex, and instances where subject incorrectly identified active tDCS as sham using a partial correlation analysis.
To examine alterations in overall L-DLPFC rsFC after active and sham tDCS, average beta-estimates for both positive and negative L-DLPFC rsFC were extracted from the subject-level COPEs images within each condition. These estimates were obtained after masking the COPEs images by the thresholded connectivity map derived from all pre-stimulation conditions. Repeated measure of analysis of variance (RM-ANOVA) and paired-sample t-tests was conducted to test the significance of difference using MATLAB. To assess locally distributed L-DLPFC rsFC changes, mixed-effect analysis was conducted using FSL FLAME1, treating participants as random effects. Pairwise post-hoc comparison was also performed.
The relationship between the E-field delivered to the L-DLPFC and changes in L-DLPFC rsFC was investigated. The positive connectivity map of active tDCST0 (cluster-extent based thresholding |z| > 3.1, FWE p < 0.05) revealed 9 clusters, while the negative connectivity map revealed 14 clusters. Average beta-estimates of 23 clusters were extracted from the subject-level COPEs images of both active tDCST0 and tDCST1.The difference in beta-estimates between active tDCST0 and tDCST1 was correlated with the E-field while controlling for age, sex, and instances where subject incorrectly identified active tDCS as sham using a partial correlation analysis.
Results:
There was a significant main effect of time on overall L-DLPFC positive and negative connectivity (F(1,20)=12.397, p=0.002, and F (1,20)=8.469, p=0.009, respectively) while the main effect of condition was not significant (F (1,20)=1.255, p=0.276 and F (1,20)=1.244, p=0.278, respectively). Post-hoc paired t-tests revealed that there was a significant decrease in rsFC after sham whereas the change was modest and not significant after active tDCS (Fig 1). No brain area was observed showing a significant main effect of condition but the left precuneus (x=-14, y=-48, z=58) exhibited an interaction effect, demonstrating that (active tDCST0 + sham tDCST1) > (active tDCST1 + sham tDCST0). A pairwise comparison revealed no significant difference between and within conditions. Among the 23 clusters showing significant connectivity with the L-DLPFC, changes in connectivity of the left DLPFC, left inferior parietal area, and bilateral lateral visual areas exhibited moderate and strong correlations with the E-field (Fig 2).
·Figure 1
·Figure 2
Conclusions:
TDCS may help to maintain overall rsFC while sham induced reduced rsFC. The impact of single-session active tDCS was subtle to change local connectivity of L-DLPFC. Nevertheless, E-field applied on the target area is associated with changes in rsFC, observed in both proximal and distally connected brain regions with L-DLPFC.
Funding: NRF-2020R1C1C1012785
Funding: NRF-2020R1C1C1012785
Brain Stimulation:
TDCS 1
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural) 2
Keywords:
FUNCTIONAL MRI
Other - transcranial direct current stimulation; dorsolateral prefrontal cortex
1|2Indicates the priority used for review