Neuronavigated Focalized Transcranial Direct Current Stimulation Administered During Functional Magnetic Resonance Imaging

Summary

This protocol describes the method of neuronavigated electrode placement for focal, transcranial direct current stimulation (tDCS) administered during functional magnetic resonance imaging (fMRI).

Abstract

Transcranial direct current stimulation (tDCS) is a noninvasive brain stimulation technique that allows the modulation of the excitability and plasticity of the human brain. Focalized tDCS setups use specific electrode arrangements to constrain the current flow to circumscribed brain regions. However, the effectiveness of focalized tDCS can be compromised by electrode positioning errors on the scalp, resulting in significant reductions of the current dose reaching the target brain regions for tDCS. Electrode placement guided by neuronavigation based on the individual's head and brain anatomy derived from structural magnetic resonance imaging (MRI) data may be suited to improve positioning accuracy.

This protocol describes the method of neuronavigated electrode placement for a focalized tDCS setup, which is suitable for concurrent administration during functional MRI (fMRI). We also quantify the accuracy of electrode placement and investigate electrode drift in a concurrent tDCS-fMRI experiment. Critical steps involve the optimization of electrode positions based on current modeling that considers the individual's head and brain anatomy, the implementation of neuronavigated electrode placement on the scalp, and the administration of optimized and focal tDCS during fMRI.

The regional precision of electrode placement is quantified using the Euclidean norm (L² Norm) to determine deviations of the actual from the intended electrode positions during a concurrent tDCS-fMRI study. Any potential displacement of electrodes (drift) during the experiment is investigated by comparing actual electrode positions before and after the fMRI acquisition. In addition, we directly compare the placement accuracy of neuronavigated tDCS to that achieved by a scalp-based targeting approach (a 10-20 Electroencephalography (EEG) system). These analyses demonstrate superior placement accuracy for neuronavigation compared to scalp-based electrode placement and negligible electrode drift across a 20 min scanning period.

Introduction

Transcranial direct current stimulation (tDCS) is a noninvasive brain stimulation technique that allows the modification of cognition and physiological brain functions in experimental and clinical contexts^1,2,3. Acute administration of tDCS can have transient changes in neuronal excitability, with the aftereffects lasting from minutes to hours after the stimulation^4,5. The applied current does not induce action potentials but rather transiently shifts the resting membrane potential of the neuron toward either de- or hyperpolarization, resulting in either increased or decreased neuronal excitability at the macroscopic level using standard protocols^4,5,6. Furthermore, regarding synaptic plasticity effects of tDCS, animal and human studies have shown that tDCS induces long-term potentiation and depression (LTP and LTD)-like processes^4,5.

In the motor system, the modulation of motor evoked potentials (MEPs) allows direct assessment of neurophysiological effects of tDCS on local cortical excitability⁷. However, this approach cannot quantify the neural effects of tDCS on higher-order cognitive functions supported by large-scale functional brain networks⁸. The effects on brain networks can be investigated by combining tDCS with modern functional imaging techniques^9,10. Among those, functional magnetic resonance imaging (fMRI) has become the most frequently used approach because it provides excellent spatial and sufficient temporal resolution to reveal the neural mechanisms by which tDCS affects local brain activity at the stimulation site and large-scale neural networks^11,12,13,14.

So far, combined fMRI-tDCS studies have mainly employed so-called conventional tDCS setups, which use relatively large rubber electrodes between 25 and 35 cm² (5 x 5 cm² and 5 x 7 cm²) inserted into saline-soaked sponge pockets^15,16. These setups project the current between two electrodes that are typically attached over (a) a target brain region for tDCS and (b) a return electrode over non-target brain regions or extracranial areas (e.g., the shoulder). This results in widespread current flow across the brain, affecting regions other than the target region, thereby complicating causal assumptions and interpretations about the neural origin of tDCS effects¹⁷.

More precise spatial targeting can be achieved by focalized tDCS¹⁸. These setups employ arrays of smaller electrodes arranged in close proximity to each other or by using a ring-shaped cathode placed around a center anode to constrain current flow to the target region^18,19. Computer simulations of electric current flow suggest that focalized tDCS can result in higher spatial precision of the current flow to the target region than conventional montages²⁰. Moreover, behavioral studies have demonstrated regional and task-specific behavioral modulation using focalized setups^19,21,22. However, only a few studies have used focalized tDCS during fMRI. These studies have been able to establish the feasibility of this approach and provided the first evidence for region-specific neural modulation^19,23.

However, because of regionally precise current delivery, focalized tDCS setups may be more sensitive to electrode positioning errors on the scalp than conventional montages. For example, Seo et al. demonstrated that positioning errors of 5 mm in a focalized motor cortex setup reduced peak somatic polarization in the hand knob by up to 87%²⁴. Moreover, a recent computational modeling study demonstrated that electrode displacement from intended positions for focal compared to conventional setups resulted in significant current dose reductions in the target regions for tDCS, ranging from 26% to 43%²⁵. Therefore, it was concluded that future studies should routinely include appropriate methods for the improvement of electrode positioning and the verification of electrode positioning before and after fMRI⁵.

In the present study, we describe the method of neuronavigated electrode positioning for a novel fMRI-compatible focal 3 x 1 tDCS setup (i.e., three individual cathodes that are arranged in a circle around a single center anode), which is currently being used in a collaborative research consortium funded by the German Science Foundation (DFG Research Unit 5429, https://www.memoslap.de). The consortium investigates behavioral and neural effects of focalized tDCS on learning and memory and predictors of stimulation response across four functional domains (i.e., visual-spatial, language, motor, and executive functions). Structural T1- and T2-weighted MRI data of study participants are acquired during a baseline scan. These data are used for individualized current flow simulations²⁶ to determine scalp positions of electrodes that maximize the current flow to the target region in individual study participants. As an example, this protocol will describe neuronavigated targeting of individually determined electrode positions centered over the right dorsolateral prefrontal cortex (rDLPFC) in one participant.

The representative results section is based on structural imaging data acquired before and after concurrent tDCS-fMRI in three subprojects of the Research Unit. These studies targeted the right occipitotemporal cortex (rOTC), left temporo-parietal cortex (lTPC), and rDLPFC. Data were acquired at the Department of Neurology of the University Medicine Greifswald. Using these data, we aimed to achieve two main objectives: (1) To quantify the spatial precision of neuronavigated electrode placement by comparing "intended" versus empirically determined "actual" electrode positions²⁵, and (2) investigate the degree of electrode displacement over the course of the fMRI sessions (i.e., electrode drift). These factors are crucial for improving the accuracy and reliability of tDCS effects in concurrent tDCS-fMRI studies²⁷. In addition, the targeting accuracy of neuronavigated tDCS is compared to that of a scalp-based approach using data from a previous tDCS-fMRI study of our group²⁵.

Protocol

All experimental procedures presented in this protocol have been reviewed and approved by the ethics committee of the University Medicine Greifswald. All participants provided informed consent prior to study inclusion and granted permission for their data to be published anonymously.

1. Screening of contraindications and general considerations

1. Prior to study enrollment, carefully screen participants for MRI²⁸ and tDCS²⁹ contraindications (e.g., pacemakers, claustrophobia, history of seizures, migraine, skin disease on the scalp [e.g., psoriasis/ eczema]) using appropriate questionnaires.
2. Explain the study goals and all planned procedures to the participants and obtain written informed consent according to the local requirements.
3. Follow general procedures for improving reporting and reproducibility of concurrent tDCS-fMRI experiments and test for potential imaging artifacts induced by the current and/or tDCS equipment as recommended by the ContES checklist³⁰.
4. Use appropriate methods to assess participant and investigator blinding^31,32 and potential adverse effects of tDCS²⁹.

2. Baseline MRI scan and individualized current modeling

1. After completing safety checks (i.e., removal of metal objects from the participant, such as coins, necklaces, piercings, etc.), guide the participant into the scanner room and position her/him comfortably on the MRI examination table. Attach the upper part of the head coil and move the participant inside the bore of the MRI scanner according to the manufacturer's specifications.
   NOTE: We used a 3T scanner equipped with a 64-channel head/neck shim coil.
2. Register the new participant using the scanner interface by clicking Main Menu | Examination | Patient Registration and fill in the required fields. Go to Program Choice and collect and select the planned imaging protocol. Click on Patient orientation and select the head first, Supine position option. From the Region of examination and laterality dropdown menu, choose brain and then click Examination to get to the examination menu.
3. Follow the onscreen instructions of the predefined scan protocol (like adjusting the field of view, etc.) to acquire T1- and T2-weighted MRI sequences. Interact with the participant via the scanner intern communication system if necessary.
   NOTE: T1- and T2-weighted images are required for individualized current modeling. The T1-weighted image is also required for neuronavigation to identify the planned and optimized positions of electrodes on the participants' scalp.
4. Use the script in https://github.com/memoslap/Greifswald and the structural imaging data acquired during the baseline MRI scan to conduct individualized current modeling (e.g., using SimNIBS²⁶). Follow the steps in the Readme.md file to apply the finite element method and individualized tetrahedral head meshes generated from the participant's structural T1- and T2-weighted images (http://simnibs.org)^26,33,34 CHARM tool³⁵ for head reconstruction to determine the peak electric field to determine optimized target positions for placement of tDCS electrodes over the rDLPFC. See Figure 1 for an example of the outcome of the procedure used for this protocol.
   NOTE: Alternative methods for identification of scalp coordinates for neuronavigated targeting are possible and depend on study-specific procedures.

3. Neuronavigation

1. Preparatory steps
   1. Turn on the Neuronavigation control computer and the tracking system.
   2. Assembly steps: Assemble all the required equipment for neuronavigation (Figure 2; for an overview of the neuronavigation setup, see Supplemental Figure S1).The equipment in Figure 2 consists of (1) a subject tracker, (2) screwdriver, (3) hexagonal rod, (4) goggles, and (5) pointer. Follow the instructions below to complete the assembly.
      1. Loosen the screw under the subject tracker with the screwdriver.
      2. Insert the longer side of the hexagonal rod into the nut mounted on the subject tracker and tighten the screw.
      3. Loosen the screw on the left side of the goggles (with the goggles positioned as if looking through them).
      4. Insert the opposite side of the hexagonal rod into the nut mounted on the left side of the goggles and tighten the screw of the goggles.
         NOTE: You have the option to install the subject tracker on the left or right side of the goggles. This choice depends on the target area relative to the position of the camera of the neuronavigation system (which must recognize the pointer and the subject tracker) in relation to the participant's position. In the present example, the tracker is attached on the left side so that the person conducting the tracking is not standing in between the tracker and the camera.
   3. Transfer the structural T1-weighted (e.g., as a Neuroimaging Informatics Technology Initiative (NIfTI) file) of the respective participant obtained in step 2.3. to the control computer of the neuronavigation system.
   4. Open the neuronavigation software and choose New Empty Project.
   5. Load the T1-weighted image of the participant and save the project by selecting save project.
   6. To initiate the 3D head reconstruction, go to the Reconstructions section in the main window of the application. Click on New… | Skin, which will open another window. Reconstruct the skin by pressing the respective button. Adjust the skin/air threshold if distortions are observed in the head reconstruction.
      NOTE: It is suggested to check if the entire head is correctly reconstructed. A good reconstruction of the nose and ears is important to detect the landmarks described in the next step.
   7. Configure five landmarks: nasion, left and right nostrils, and left and right preauricular pits (LPA and RPA). These are required to register the participant in the Landmarks section of the main window of the application (for details, see Supplemental Figure S2).
   8. Configure the electrode position for the rDLPFC in the Targets section by inserting the x, y, and z coordinates of the anode position (provided by the individualized current flow simulations described in step 2.4), click on Add new, and type Anode as the name of the electrode position. Repeat the procedure for the three return electrodes.
2. Neuronavigated identification of electrode positions
   1. Position the participant comfortably in a chair facing the tracking camera. Ask him/her to put on the goggles with the subject tracker attached.
   2. Instruct participants not to touch the goggles during the entire neuronavigation procedure. This is crucial for a precise registration of the landmarks and validation of the electrode positions.
   3. Go to the Sessions tab, and from the left corner, choose Online Session from the New dropdown menu. Select the Polaris tab and verify the visibility of the subject tracker and the pointer by moving them into the camera field of view. Both devices are correctly recognized when the corresponding red crosses change to the green check marks in the Tools panel (left side of the application window).
   4. Select the Registration section to register the five predefined landmarks. Find the respective landmark by placing the pointer perpendicularly to the participant's head with the sensors pointing to the camera. Then, press the foot pedal of the neuronavigation system to confirm the position. Ensure that a green checkmark is seen in front of the name of each landmark.
   5. Navigate to the validation section and validate the landmarks by placing the tip of the pointer on the registered landmarks and check the two distance indices; the first index shows the distance between the crosshair (virtual pointer tip) and the registered landmark, the second index shows the distance between the crosshair and the reconstructed skin.
      ​NOTE: When using the default parameters of the neuronavigation system, indices below 5 mm indicate sufficient precision to pass the validation step. This is acceptable for many experimental contexts. However, because of the optimized focal setup used in this protocol, the validation of electrodes is only accepted if the deviation from the intended coordinates is less than 1 mm (see Figure 3).
   6. Move the pointer over the participant's scalp and check the crosshairs of the reconstructed scalp on the neuronavigation monitor. If the crosshairs remain aligned with the reconstructed scalp without penetrating it or creating a gap above it, it is acceptable, although not recommended, to skip the next step.
   7. Sample additional points around extreme point locations, including the left-most, right-most, top-most, back-most, and front-most locations. To do so, click the Add button in the refinement landmarks panel for each position. Next, position the pointer on the target surface of the head, ensuring the tip of the pointer gently touches the scalp, and press the pedal to register the position. Repeat this process until the distance between the crosshair and the reconstructed skin is as low as possible.
      NOTE: Different neuronavigation systems may use different terminology to refer to this process (e.g., surface registration).
   8. Select the Perform section and move the pointer to the approximate location of the DLPFC to find the position of the center anode (based on the coordinates provided in Figure 1D). While moving the pointer, observe the screen simultaneously. Mark the positions of the electrodes when the tip of the pointer aligns with the center of the green crosshair on the screen.
   9. Move the participant's hair away from the corresponding area on the scalp and mark the positions with a skin marker/pen. Repeat the process for the positions of the three cathodes (see Figure 1D).
   10. Apply a small amount of topical anesthetic cream at the intended positions of the electrodes to reduce the physical sensations on the scalp during tDCS-fMRI.
       NOTE: Ensure at least 20 min pass between the application of the cream and starting the tDCS.

4. tDCS-fMRI

1. Preparation for concurrent tDCS-fMRI
   1. Prepare focal 3 x 1 tDCS with a multi-channel direct current (DC) stimulator.
   2. Use the multi-channel DC stimulator³⁶ in normal (non-battery) mode. Insert the power plug of the stimulator into the same power strip as the scanner for enhanced signal-to-noise ratio and reduction of imaging artifacts (see also notes from the manufacturer's specifications).
   3. Make sure all necessary stimulation-related materials are available and clean (Figure 4). Take special care that the rubber electrodes and 3D templates do not contain any paste from previous experimental sessions.
      NOTE: The setup used in this project employs customized electrodes and filter boxes developed in collaboration with the manufacturer of the DC stimulator to meet the specific requirements of the Research Unit (see Figure 4A). However, standard MRI-compatible components can also be used.
      1. Use a 3D-printed (thermoplastic) electrode paste fill aid to standardize the application of conductive electrode paste prior to attaching the electrodes to the scalp (Figure 4A (9),B).
      2. Use a 3D-printed (thermoplastic) spacer to position the electrodes on the scalp and ensure that the distances between the anode and the cathodes are maintained during the fMRI-tDCS sessions (Figure 4A (10),C).
         NOTE: The 3D customized templates can be accessed using the following link: https://github.com/memoslap/Material
   4. To set up the DC stimulator, connect the DC stimulator to the outer box (by using the outer box cable and adapter). Connect the inner box cable to the inner and outer box (see Figure 4).
   5. Apply 1 mm of conductive paste evenly on the surface of all electrodes of the focal-tDCS 3x1 setup. Use the electrode fill aid to standardize the paste thickness. Cover only the surface of the electrode with the paste and remove any additional paste.
   6. Turn on the DC stimulator and the Panel PC (in the given order). Double-click on the DC-Stimulator MC icon on the desktop. Select the required stimulation sequence at the selected sequence setup dropdown menu. Click on the calibrate stimulator button to calibrate the stimulator with no electric load (i.e., the participant is not connected to the stimulator).
      CAUTION: If the participant is connected to the stimulator while calibrating, the electric current can induce painful sensations.
   7. Position the participant comfortably near the DC stimulator outside the MRI scanner room.
   8. Determine the widest part of the participant's head, from the forehead (just above the eyebrows) to the occipital bone at the back of the head, to choose the optimal size of the EEG cap to keep the electrodes in place during the tDCS-fMRI session.
   9. Place the electrodes in the spacer to ensure equal spacing of the cathodes around the center anode and attach the electrodes over the identified and marked scalp positions.
   10. Use the optimal size EEG cap without plastic inserts to keep the electrodes in place during tDCS-fMRI.
       NOTE: Make sure that electrodes do not get displaced while the cap is put in place.
   11. Connect the leads of the electrode cables to the inner box of the DC stimulator to perform an impedance check. Select the required stimulation sequence at the selected sequence setup dropdown menu. Initiate the impedance check by pressing the impedance check button on the DC stimulator. Make sure that MR mode is checked.
   12. Conduct the Impedance Check by pressing the respective button on the stimulator interface; if the impedance is ≤25 kΩ, go to the next step (4.1.13). If the impedance is higher for any electrode, softly press the electrodes to the scalp, tighten the cap, and allow the conductive paste to warm up. If required, apply more paste.
   13. Disconnect the inner box from the outer box and insert the outer box into the waveguide of the scanner.
   14. Guide the participant into the scanner room while holding on to the inner box and connected electrode cables.
       NOTE: At any time, take great care that there is no tension on the cables, which could result in electrode displacement.
   15. Ask the participant to sit on the MRI examination table and re-connect the inner box to the outer box (that is inserted in the waveguide of the scanner).
   16. Place the participant comfortably in a supine position on the MRI examination table, with the head positioned in the open head coil. Use inflatable cushions on both sides of the head and an extra cushion at the top of the head to stabilize it.
       ​NOTE: Inflatable cushions are preferred at the sides to foam cushions to avoid electrode displacement when the foam cushions are inserted.
   17. Lead the electrode cables through the lower part of the head coil before attaching the upper part of the head coil and locking it in place.
   18. Position the inner box next to the participant on the MRI examination table and move the participant into the scanner bore.
   19. Leave the scanner room and inform the participant about the upcoming procedures via the communication interface of the scanner prior to each structural and functional imaging sequence, as well as prior to the second impedance check inside the scanner.
2. Concurrent tDCS-fMRI
   1. On the scanner Panel PC, register the new participant by clicking Main Menu | Examination | Patient Registration and fill in the required necessary fields. Go to Program Choice and collect and select the planned imaging protocol. Click on Patient orientation and select the head first, Supine position option. From the Region of examination and laterality dropdown menu, choose brain and click Examination to get to the examination menu.
   2. Follow the onscreen instructions to acquire the planned scans (pre-fMRI Pointwise Encoding Time reduction with Radial Acquisition (PETRA), fMRI, post-fMRI PETRA) in the order shown in the following steps of this protocol.
      1. Acquire a PETRA scan that allows verification of the electrode positions on the participant's head.
      2. Inform the participant that two 10 min resting-state fMRI scans will be conducted and that he/she should maintain their gaze on a fixation cross (displayed via a projector and mirrors mounted on the head coil) for the entire duration of the scanning period (2 x 10 min).
      3. Start the tDCS stimulation by pressing the init stimulation button on the Panel PC. Click the release start-trigger button to start the stimulation with a 10 s ramp-up prior to commencing with the functional imaging sequences. Administer tDCS for 20 min with 2 mA.
         NOTE: Other stimulation protocols with different ramping periods or stimulation intensities or durations are possible.
      4. After the end of the functional scans and stimulation period, acquire a second PETRA scan while the electrodes are still attached to the participant's head.
         NOTE: By comparison with the pre-fMRI PETRA scan, this allows the determination of potential electrode movement across the tDCS-fMRI experiment (i.e., drift).
   3. Upon completion of the MRI session, disconnect the electrode cables from the outer box and turn off the DC stimulator, move the participant out of the scanner bore, remove the cap from the participant's head, and remove the electrodes.
   4. Inspect the participant's scalp for potential skin redness caused by the stimulation. Clean the participant's scalp where the electrodes were placed.
   5. At the end of the experiment, ask the participant to complete the tDCS adverse effects²⁹ and blinding^31,32 questionnaires.

Results

Data from 43 healthy young participants (20 men/23 women, aged 24.74 ± 5.50 years) were included. The participants completed up to four fMRI sessions. Neuronavigated placement of electrodes was conducted prior to each fMRI session. In total, 338 datasets representing the positions of the center anodes before and after fMRI were included in data analyses.

To determine the intended positions of the electrodes, individualized current modeling was performed using structural M...

Discussion

Critical steps, potential modifications, and troubleshooting of the method
Accurate positioning of electrodes is a crucial technical factor in tDCS experiments, and deviations from intended scalp positions or electrode drift can affect current flow to the intended target brain regions^42,43. This is particularly relevant for focalized tDCS, as the regional specificity of the administered current makes these setups particularly susceptible to...

Materials

Name	Company	Catalog Number	Comments
Brainsight neuronavigation system	Brainsight; Rogue Research Inc., Montréal, Canada
CR-5 Pro high temp 3D printer	CREALITY, Shenzhen, China
DC-STIMULATOR MC	NeuroConn GmbH, Ilmenau, Germany	https://www.neurocaregroup.com/technology/dc-stimulator-mc
EMLA Cream 5%	Aspen, Dublin, Ireland
MAGNETOM Vida 3T, syngo_MR_XA50 software	Siemens Healthineers AG, Forchheim, Germany
Polaris camera	Polaris Vicra; Northern Digital Inc., Waterloo, Canada
Ten20 conductive EEG paste	Weaver and Company, Aurora, USA
TPU 3D printer filament	SUNLU International, Hong-Kong, China
Example of alternatives
Ingenia 3.0T (MR-scanner)	Phillips, Amsterdam, Netherlands
Localite TMS Navigator (Neuronavigation equipment)	Localite, Bonn, Germany
Neural Navigator (Neuronavigation equipment)	Soterix, New Jersey, USA
PEBA 3D printer filament	Kimya, Nantes, France
PLA 3D printer filament	Filamentworld, Neu-Ulm, Deutschland
StarStim (Stimulator)	Neuroelectrics, Barcelona, Spain