Transcranial direct current stimulation with functional magnetic resonance imaging: a detailed validation and operational guide

Abstract

Introduction: Transcranial direct current stimulation (tDCS) is a non-invasive brain stimulation technique used to modulate human brain and behavioural function in both research and clinical interventions. The combination of functional magnetic resonance imaging (fMRI) with tDCS enables researchers to directly test causal contributions of stimulated brain regions, answering questions about the physiology and neural mechanisms underlying behaviour. Despite the promise of the technique, advances have been hampered by technical challenges and methodological variability between studies, confounding comparability/replicability. Methods: Here tDCS-fMRI at 3T was developed for a series of experiments investigating language recovery after stroke. To validate the method, one healthy volunteer completed an fMRI paradigm with three conditions: (i) No-tDCS, (ii) Sham-tDCS, (iii) 2mA Anodal-tDCS. MR data were analysed in SPM12 with region-of-interest (ROI) analyses of the two electrodes and reference sites. Results: Quality assessment indicated no visible signal dropouts or distortions introduced by the tDCS equipment. After modelling scanner drift, motion-related variance, and temporal autocorrelation, we found no field inhomogeneity in functional sensitivity metrics across conditions in grey matter and in the three ROIs. Discussion: Key safety factors and risk mitigation strategies that must be taken into consideration when integrating tDCS into an fMRI environment are outlined. To obtain reliable results, we provide practical solutions to technical challenges and complications of the method. It is hoped that sharing these data and SOP will promote methodological replication in future studies, enhancing the quality of tDCS-fMRI application, and improve the reliability of scientific results in this field. Conclusions: The method and data provided here provide a technically safe, reliable tDCS-fMRI procedure to obtain high quality MR data. The detailed framework of the Standard Operation Procedure SOP (https://doi.org/10.5281/zenodo.4606564) systematically reports the technical and procedural elements of our tDCS-fMRI approach, which we hope can be adopted and prove useful in future studies.

Transcranial electrical stimulation effects on neurovascular coupling

“…The poles and the damping parameters associated with the linearized models of the four tES perturbation pathways are listed in Table S1 in the supplementary materials. S1 in the supplementary materials show the electrical field distribution for 2mA tDCS with FC5 (anodal electrode) and FP2 (cathodal electrode) computed with the ROAST package [14] for the opensource tES-fMRI dataset [10]. Figure 5(a) shows the functional sensitivity metrics calculated using the open-source code and data from the tES-fMRI dataset [10], where anodal tDCS (figure legend: anodal) led to a shift to a higher t-score (first row of Figure 5 We also applied HRF estimation using the rsHRF toolbox [13] that elucidated the effects of anodal tDCS at four ROIs in the grey matter.…”

“…S1 in the supplementary materials show the electrical field distribution for 2mA tDCS with FC5 (anodal electrode) and FP2 (cathodal electrode) computed with the ROAST package [14] for the opensource tES-fMRI dataset [10]. Figure 5(a) shows the functional sensitivity metrics calculated using the open-source code and data from the tES-fMRI dataset [10], where anodal tDCS (figure legend: anodal) led to a shift to a higher t-score (first row of Figure 5 We also applied HRF estimation using the rsHRF toolbox [13] that elucidated the effects of anodal tDCS at four ROIs in the grey matter. Figure 5(b) shows that the canonical HRF with time & dispersion derivatives primarily captured the tES effects on the magnitude of the main response and the magnitude of undershoot.…”

“…In this study, we used the modal analysis approach to analyze the physiologically detailed NVU model for evaluating neurovascular coupling modes that may be perturbed with tACS. We also used a nonparametric impulse response estimation of the HRF from an open-source fMRI-tDCS dataset [10] and compared that with the canonical HRF to elucidate tES effects on the temporal profile of the HRF.…”

“…In contrast, in transcranial alternating current stimulation (tACS), the oscillating current reverses the flow rhythmically at a specific frequency. For example, tACS differs from tDCS in that it provides a mechanism for manipulating intrinsic oscillations through the injection of sinusoidal currents [8][9][10]. Then, the other methods are transcranial oscillating current stimulation (tOCS) which uses tDCS to set a baseline to the tACS oscillations, and transcranial random noise stimulation (tRNS) that injects 'noisy' current with bounded stochasticity [7,8].…”

“…FigureS1in the supplementary materials show the electrical field distribution for 2mA tDCS with FC5 (anodal electrode) and FP2 (cathodal electrode) computed with the ROAST package[14] for the opensource tES-fMRI dataset[10]. Figure5(a) shows the functional sensitivity metrics calculated using the open-source code and data from the tES-fMRI dataset[10], where anodal tDCS (figure legend: anodal) led to a shift to a higher t-score (first row of Figure5(a)). Here, the No-tDCS and the Sham-tDCS conditions resulted in a similar HRF temporal profile different from the anodal tDCS condition based on the canonical HRF model.…”