Presented here is a protocol to achieve higher accuracy in determination of stimulation location combining a 3D digitizer with high-definition transcranial direct current stimulation.
The abundance of neuroimaging data and rapid development of machine learning has made it possible to investigate brain activation patterns. However, causal evidence of brain area activation leading to a behavior is often left missing. Transcranial direct current stimulation (tDCS), which can temporarily alter brain cortical excitability and activity, is a noninvasive neurophysiological tool used to study causal relationships in the human brain. High-definition transcranial direct current stimulation (HD-tDCS) is a noninvasive brain stimulation (NIBS) technique that produces a more focal current compared to conventional tDCS. Traditionally, the stimulation location has been roughly determined through the 10-20 EEG system, because determining precise stimulation points can be difficult. This protocol uses a 3D digitizer with HD-tDCS to increase accuracy in determination of stimulation points. The method is demonstrated using a 3D digitizer for more accurate localization of stimulation points in the right temporo-parietal junction (rTPJ).
Transcranial direct current stimulation (tDCS) is a noninvasive technique that modulates cortical excitability with weak direct currents over the scalp. It aims to establish causality between neural excitability and behavior in healthy humans1,2,3. In addition, as a motor neurorehabilitation tool, tDCS is widely used in the treatment of Parkinson’s disease, stroke, and cerebral palsy4. Existing evidence suggests that traditional pad-based tDCS produces current flow through a relatively larger brain region5,6,7. High-definition transcranial direct current stimulation (HD-tDCS), with the center ring electrode sitting over a target cortical region surrounded by four return electrodes8,9, increases focality by circumscribing four ring areas5,10. In addition, changes in excitability of the brain induced by HD-tDCS have significantly larger magnitudes and longer durations than those generated by traditional tDCS7,11. Therefore, HD-tDCS is widely used in research7,11.
Noninvasive brain stimulation (NIBS) requires specialized methods to ensure that a stimulation site is present in the standard MNI and Talairach systems12. Neuronavigation is a technique that allows for mapping interactions between transcranial stimuli and the human brain. Its visualization and 3D image data are used for precise stimulation. In both tDCS and HD-tDCS, a common assessment of stimulation sites on the scalp is typically the EEG 10-20 system13,14. This measurement is widely used for placing the tDCS pads and optode holders for functional near infrared spectroscopy (fNIRS) in the initial stage13,14,15.
Determining the precise stimulation points when using the 10-20 system can be difficult (e.g., in the temporo-parietal junction [TPJ]). The best way to solve this is to obtain structural images from participants using magnetic resonance imaging (MRI), then obtain the exact probe position by matching target points to their structural images using digitizing products15. MRI provides good spatial resolution but is expensive to use15,16,17. Moreover, some participants (e.g., those with metal implants, claustrophobic people, pregnant women, etc.) cannot be subjected to MRI scanners. Therefore, there is a strong need for a convenient and efficient way to overcome the abovementioned limitations and increase accuracy in determining stimulation points.
This protocol uses a 3D digitizer to overcome these limitations. Compared to MRI, key advantages of a 3D digitizer are low costs, simple application, and portability. It combines five reference points (i.e., Cz, Fpz, Oz, left preauricular point, and right preauricular point) of individuals with location information of the target stimulation points. Then, it produces a 3D position of electrodes on the subject’s head and estimates their cortical positions by fitting with the vast data from the structural image12,15. This probabilistic registration method enables the presentation of transcranial mapping data in the MNI coordinate system without recording a subject’s magnetic resonance images. The approach generates anatomical automatic labels and Brodmann areas11.
The 3D digitizer, used to mark space coordinates based on the data from structural images, was first used to determine the position of optodes in fNIRS research18. For those who use HD-tDCS, a 3D digitizer breaks the finite stimulation points of the EEG 10-20 system. The distance of the four return electrodes and center electrode is flexible and can be adjusted as needed. When using the 3D digitizer with this protocol, the coordinates of the rTPJ were obtained, which is beyond the 10-20 system. Also shown are the procedures for targeting and stimulating the right temporo-parietal junction (rTPJ) of the human brain.
The protocol meets the guidelines of the Institutional Review Board of Southwest University.
1. Determination of Stimulation Location
2. Preparation of Electrode Holding Cap
NOTE: The following steps are shown in Figure 1.
3. 3D Digitizer Measurement
4. Data Conversion and Spatial Registration
5. Stimulation
6. Post-stimulation
Using the methods presented, coordinates of the rTPJ were determined, which requires stimulation points beyond the 10-20 system. First, the circumference of the headform should be similar to the actual head. Here, the length of the nasion to inion of the headform was ~36 cm, and the length between the bilateral preauricular was ~37 cm.
The steps for producing the electrode cap guide the measuring positions of the 10-20 system. Here, Nz, Iz, Cz, Fpz, Oz, Pz, T8, T7, C4, P8, O2, P4, C6, P6, and CP6 were determined. The approximate location of the RTPJ (about the midpoint between CP6 and P6) was found on the scalp. The distance between the central and peripheral electrodes should be adjusted based on experimental objectives. Previous research obtained radius values ranging from 3.5–7.5 cm11,14,30. With different radius values, DC intensity and stimulation duration may generate different electric field strengths. In this protocol, the distance between all return electrodes and the central active electrode were fixed to 3.5 cm.
Several important reference points on the swimming cap were kept, including Fpz, Cz, Oz, T8, and C4. The Vertex on the scalp was located before the stimulation, and it is critical that the Cz point on the cap exactly aligns with the Vertex. Once the cap is in position, the cap should not move. One .mat file and two .csv files after digitizing were obtained (i.e., sub01_origin.csv, which included the coordinate information of the reference [with subject number 01]), while sub01_others.csv included the coordinate information of the five targeted points [with subject number 01)].
Three .txt files were obtained after data conversion and spatial registration. In digitizer software, there are transmitter, detector (receiver), and channel options for fulfilling the requirements of fNIRS experiments. The coordinate data of the transmitter, detector, or channel should be the same. However, small operating errors may occur, because of laboratory personnel skills, pen holding gesture, etc.
Using the NIRS-SPM stand-alone registration function, the spatial registration function generates MNI coordinates. The numbers in the first line in Table 1 represent the order in the digitizer. In this protocol, the data from number five is the position information about the center electrode. In Brodmann areas (BA), the anatomical label and its number were obtained. The number after each line indicates the percentage of overlap. In anatomical automatic labels (AAL), the anatomical label and percentage of overlap were obtained. To reduce measurement errors, the average value of three data points from the five electrodes’ final MNI coordinates were calculated. As for AAL and BA, the value represents a percentage of overlap with the cerebral cortex. All possibilities were combined into final data (Table 1).
According to the data from MNI coordinates, AAL, and BA, if the difference between the value and target value is too large, the swimming cap must be adjusted to the relative position of the actual values of X, Y, Z, and the target value, as explained in sections 2–411,14,30,31.
Figure 1: Steps for creating the holding electrode cap. Please click here to view a larger version of this figure.
Figure 2: 3D digitizer. The 3D digitizer is a cost-effective solution for 3D digitizing. It is a dual sensor motion tracker. The source is a magnetic transmitter that emits an electromagnetic dipole field. The sensor is a receiver that detects the field. The stylus allows accurate pinpointing of X, Y, and Z data points. The control box connects to the computer and transfer data. Please click here to view a larger version of this figure.
Figure 3: Necessary materials for stimulation. These materials include a tDCS device, 4×1 Multichannel Stimulation Adapter, four 9 V batteries, five Ag/AgCI sodium ring electrodes, five HD plastic casings and their respective caps, electrically conductive gel, a syringe, a standard tape measure, and a swimming cap. Please click here to view a larger version of this figure.
1 | 2 | 3 | 4 | 5 | ||||||||||||
MNI | X | Y | Z | X | Y | Z | X | Y | Z | X | Y | Z | X | Y | Z | |
Channel | 43 | -89 | 13 | 46 | -64 | 54 | 71 | -29 | 25 | 64 | -56 | -16 | 60 | -66 | 24 | |
Transmit | 42 | -89 | 18 | 42 | -67 | 55 | 71 | -32 | 27 | 64 | -57 | -16 | 60 | -66 | 24 | |
Receiver | 43 | -89 | 16 | 45 | -67 | 54 | 71 | -31 | 27 | 65 | -58 | -12 | 58 | -69 | 22 | |
Mean | 42.7 | -89 | 15.7 | 44.3 | -66 | 54.3 | 71 | -30.7 | 26.3 | 64.3 | -57 | -14.7 | 59.3 | -67 | 23.3 | |
BA | Channel | 18 – Visual Association Cortex (V2), 0.27823 | 7-Somatosensory Association Cortex, 0.27876 | 2 –Primary Somatosensory Cortex, 0.41667 | 20 – Inferior Temporal gyrus, 0.089606 | 21 – Middle Temporal gyrus, 0.0072464 | ||||||||||
19 – V3, 0.72177 | 39 – Angular gyrus, part of Wernicke's area, 0.53982 | 22 – Superior Temporal Gyrus, 0.28086 | 37 – Fusiform gyrus, 0.91039 | 22 – Superior Temporal Gyrus, 0.17391 | ||||||||||||
40 – Supramarginal gyrus part of Wernicke's area, 0.18142 | 40 – Supramarginal gyrus part of Wernicke's area, 0.19136 | 37 – Fusiform gyrus, 0.07971 | ||||||||||||||
48 – Retrosubicular area, 0.11111 | 39 – Angular gyrus, part of Wernicke's area, 0.73913 | |||||||||||||||
Transmit | 18 – Visual Association Cortex (V2), 0.15936 | 7 – Somatosensory Association Cortex, 0.57466 | 2 – Primary Somatosensory Cortex, 0.38871 | 20 – Inferior Temporal gyrus, 0.035842 | 21 – Middle Temporal gyrus, 0.0072464 | |||||||||||
19 – V3, 0.84064 | 39 – Angular gyrus, part of Wernicke's area, 0.34389 | 22 – Superior Temporal Gyrus, 0.15674 | 37 – Fusiform gyrus, 0.96416 | 22 – Superior Temporal Gyrus, 0.17391 | ||||||||||||
40 – Supramarginal gyrus part of Wernicke's area, 0.081448 | 40 – Supramarginal gyrus part of Wernicke's area, 0.31034 | 37 – Fusiform gyrus, 0.07971 | ||||||||||||||
48 – Retrosubicular area, 0.1442 | 39 – Angular gyrus, part of Wernicke's area, 0.73913 | |||||||||||||||
Receiver | 18 – Visual Association Cortex (V2), 0.21514 | 7 – Somatosensory Association Cortex, 0.42601 | 2 – Primary Somatosensory Cortex, 0.44025 | 20 – Inferior Temporal gyrus, 0.0071429 | 19 – V3, 0.0036101 | |||||||||||
19 – V3, 0.78486 | 39 – Angular gyrus, part of Wernicke's area, 0.51121 | 22 – Superior Temporal Gyrus, 0.14151 | 37 – Fusiform gyrus, 0.99286 | 22 – Superior Temporal Gyrus, 0.054152 | ||||||||||||
40 – Supramarginal gyrus part of Wernicke's area, 0.06278 | 40 – Supramarginal gyrus part of Wernicke's area, 0.28302 | 37 – Fusiform gyrus, 0.12274 | ||||||||||||||
48 – Retrosubicular area, 0.13522 | 39 – Angular gyrus, part of Wernicke's area, 0.81949 | |||||||||||||||
AAL | Channel | Occipital_Mid_R, 1 | Parietal_Sup_R, 0.030973 | SupraMarginal_R, 0.65741 | Temporal_Mid_R, 0.039427 | Occipital_Mid_R, 0.13406 | ||||||||||
Parietal_Inf_R, 0.31416 Angular_R, 0.65487 | Temporal_Sup_R, 0.34259 | Temporal_Inf_R, 0.93907 | Angular_R, 0.33696 | |||||||||||||
Cerebelum_Crus1_R,0.021505 | Temporal_Sup_R,0.032609 | |||||||||||||||
Temporal_Mid_R, 0.49638 | ||||||||||||||||
Transmit | Occipital_Mid_R, 1 | Parietal_Sup_R, 0.20814 | SupraMarginal_R, 0.74922 | Temporal_Mid_R, 0.032258 | Occipital_Mid_R, 0.13406 | |||||||||||
Parietal_Inf_R, 0.20362 | Temporal_Sup_R, 0.25078 | Temporal_Inf_R, 0.94265 | Angular_R, 0.33696 | |||||||||||||
Angular_R, 0.58824 | Cerebelum_Crus1_R, 0.02509 | Temporal_Sup_R,0.032609 | ||||||||||||||
Temporal_Mid_R, 0.49638 | ||||||||||||||||
Receiver | Occipital_Mid_R, 1 | Parietal_Sup_R, 0.044843 | SupraMarginal_R, 0.7673 | Temporal_Mid_R, 0.11429 | Occipital_Mid_R, 0.22022 | |||||||||||
Parietal_Inf_R, 0.20179 | Temporal_Sup_R, 0.2327 | Temporal_Inf_R, 0.88571 | Angular_R, 0.15523 | |||||||||||||
Angular_R, 0.75336 | Temporal_Mid_R, 0.62455 |
Table 1: Localization of stimulations in the brain area. Please click here to view this table (Right click to download).
Supplementary File. Please click here to view this file (Right click to download).
Compared to traditional tDCS, HD-tDCS increases the focality of stimulation. Typical sites of stimulation are often based on the 10-20 EEG system. However, determining the precise stimulation points beyond this system can be difficult. This paper combines a 3D digitizer with HD-tDCS to determine stimulation points beyond the 10-20 system. It is important to clearly define the steps and precautions for making and using the electrode cap in such cases.
In general, the position of target stimulation areas is derived from the results of previous brain imaging studies, and the position of the stimulation areas on 10-20 international system or MNI coordinates can be obtained. The steps for creating the electrode cap guide for measuring positions of the 10-20 system are critical. It is key that the reference on the cap aligns with the international 10-20 system for scalp locations when placing the cap on the head. Once the 3D digitizer starts running, the source and sensor should not move, or it will cause data deviation.
In the software, the reference points are on the scalp and not on the cap, unless all the reference points of scalp and cap are matching. If the error between the measured results and target values is out of the acceptable range, the position of the marked points should be slightly adjusted. After adjustment, measurements should then be made again. Once users press the “MODE SELECT” button and switch from “SCAN” to “PASS”, the current will start passing from the conventional tDCS device through the electrodes into the 4×1 Multichannel Stimulation Adapter.
The modular electroencephalogram recording cap provides fixed positions of probes. However, determining the precise stimulation points beyond this system can be difficult. The positions of electrodes beyond the 10-20 system can be determined using the protocol described, as well as the coordinates of stimulation points. The radius setting should be based on the experimental objectives. Using the method described here, the radius of the four return electrodes and center electrode can be flexibly adjusted.
There are many digitizer software packages (e.g., the Brainstorm software for an fNIRS task; here, the Vpen software was used)15. Different data collection software packages emphasize different functions and should be selected according to the research question. Head circumference varies among individuals; hence, using the same cap can produce errors. However, the modular electroencephalogram recording cap also suffers from this problem.
The authors have nothing to disclose.
This study was supported by the National Natural Science Foundation of China (31972906), Entrepreneurship and Innovation Program for Chongqing Overseas Returned Scholars (cx2017049), Fundamental Research Funds for Central Universities (SWU1809003), Open Research Fund of the Key Laboratory of Mental Health, Institute of Psychology, Chinese Academy of Sciences (KLMH2019K05), Research Innovation Projects of Graduate Student in Chongqing (CYS19117), and the Research Program Funds of the Collaborative Innovation Center of Assessment toward Basic Education Quality at Beijing Normal University (2016-06-014-BZK01, SCSM-2016A2-15003, and JCXQ-C-LA-1). We would like to thank Prof. Ofir Turel for his suggestions on the early draft of this manuscript.
1X1 Low Intensity transcranial DC Stimulator | Soterix Medical | 1300A | |
3-dimensional Polhemus-Patriot Digitizer | POLHEMUS | 1A0453-001 | PATRIOT system component |
4X1 Multi-Channel Stimulation Interface | Soterix Medical | 4X1-C3 | |
Dell desktop computer | Dell | CRFC4J2 | Master computer to run 3D digitizer application |