NeuroImage 87 (2014) 332–344

Contents lists available at ScienceDirect

NeuroImage journal homepage: www.elsevier.com/locate/ynimg

A computational modelling study of transcranial direct current stimulation montages used in depression Siwei Bai a, Socrates Dokos a, Kerrie-Anne Ho b,d, Colleen Loo b,c,d,⁎ a

Graduate School of Biomedical Engineering, Faculty of Engineering, University of New South Wales (UNSW), NSW 2052, Australia School of Psychiatry, UNSW, NSW 2052, Australia Department of Psychiatry, St George Hospital, NSW 2217, Australia d Black Dog Institute, NSW 2031, Australia b c

a r t i c l e

i n f o

Article history: Accepted 5 November 2013 Available online 15 November 2013 Keywords: Transcranial direct current stimulation Depression Computational model

a b s t r a c t Transcranial direct current stimulation (tDCS) is a neuromodulatory technique which involves passing a mild electric current to the brain through electrodes placed on the scalp. Several clinical studies suggest that tDCS may have clinically meaningful efficacy in the treatment of depression. The objective of this study was to simulate and compare the effects of several tDCS montages either used in clinical trials or proposed, for the treatment of depression, in different high-resolution anatomically-accurate head models. Detailed segmented finite element head models of two subjects were presented, and a total of eleven tDCS electrode montages were simulated. Sensitivity analysis on the effects of changing the size of the anode, rotating both electrodes and displacing the anode was also conducted on selected montages. The F3–F8 and F3–F4 montages have been used in clinical trials reporting significant antidepressant effects and both result in relatively high electric fields in dorsolateral prefrontal cortices. Other montages using a fronto-extracephalic or fronto-occipital approach result in greater stimulation of central structures (e.g. anterior cingulate cortex) which may be advantageous in treating depression, but their efficacy has yet to be tested in randomised controlled trials. Results from sensitivity analysis suggest that electrode position and size may be adjusted slightly to accommodate other priorities, such as skin discomfort and damage. © 2013 Elsevier Inc. All rights reserved.

Introduction Transcranial direct current stimulation (tDCS) is a neuromodulatory technique which involves passing a mild electric current to the brain through electrodes placed on the scalp. This direct constant flow of current modulates underlying cortical activity with specific outcomes related to anodal or cathodal stimulation (Nitsche and Paulus, 2000, 2001). The relative position (electrode montage) and size of the anode and cathode determine the distribution of current density throughout the brain (Bikson et al., 2010; Datta et al., 2011; Lee et al., 2012; Miranda et al., 2009; Wagner et al., 2007). Thus there is potential for stimulation to be focussed on specific cortical brain regions for therapeutic or investigative purposes or more diffuse effects can be produced if widespread activation of brain regions is desired. A key application of tDCS has been investigated in the treatment of depression. Several recent open label and placebo-controlled trials, and a meta-analysis of mean change in depression scores from placebo-controlled studies suggest that tDCS may have clinically ⁎ Corresponding author at: Black Dog Institute, Hospital Road, Prince of Wales Hospital, Randwick, NSW 2031, Australia. Fax: +61 2 9113 3734. E-mail addresses: [email protected] (S. Bai), [email protected] (S. Dokos), [email protected] (K.-A. Ho), [email protected] (C. Loo). 1053-8119/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.neuroimage.2013.11.015

meaningful efficacy (Boggio et al., 2008; Brunoni et al., 2011; Fregni et al., 2006a, 2006b; Kalu et al., 2012; Loo et al., 2012; Martin et al., 2011; Palm et al., 2011). These studies focused on anodal stimulation of the left dorsolateral prefrontal cortex (DLPFC), based on observations that this area has been associated with underactivity in depression (Grimm et al., 2008). However, studies differed in the location of the cathode, i.e. the return electrode — right supraorbital, right lateral orbitofrontal, right DLPFC or in an extracephalic position. Though the anodal left DLPFC electrode is often considered the “active” electrode, the placement of the cathode is important for several reasons: shunting of much of the current over the scalp may occur if the inter-electrode distance is too close (Datta et al., 2008; Miranda et al., 2006; Weaver et al., 1976), current density under the anode is affected by the placement of the reference or “return” electrode (Bikson et al., 2010; Datta et al., 2011), and the pattern of brain areas stimulated will be determined by the overall montage. All of these factors may have important therapeutic implications. Pathophysiological changes in depression are system-wide, involving a network of various cortical and limbic structures rather than a solitary brain region such as the left DLPFC (Mayberg, 2007). Hypoactivity in cortical regions and hyperactivity in subcortical and limbic regions is often associated with symptoms of depression (Fitzgerald et al., 2008; Mayberg, 1997). Meta-analyses have identified frontal and temporal

S. Bai et al. / NeuroImage 87 (2014) 332–344

cortices, the insula and cerebellum as regions of hypoactivity while subcortical and limbic regions tend to be hyperactive. This distributed network of structures includes the DLPFC, medial prefrontal cortex (MPFC), orbitofrontal cortex (OFC), as well as the anterior cingulate cortex (ACC), insula and hippocampus (Fox et al., 2012; Mayberg, 2003). Most recently, functional connectivity studies have suggested altered activity at a network level during the resting state (Carballedo et al., 2011). In particular, there is increased functional connectivity in the subgenual anterior cingulate (sgACC), thalamus and OFC in people with depression (Greicius et al., 2007). Further, overactivity in the sgACC has been shown to be strongly negatively correlated with resting state underactivity in the left DLPFC (Fox et al., 2012). Studies of deep brain stimulation (DBS) in depression have also provided insight into the critical regions involved in depression. Consistent with imaging studies, DBS interventions targeted at the sgACC have demonstrated efficacy in reducing symptoms of depression (Lozano et al., 2012; Mayberg et al., 2005). DBS to specific regions of the basal ganglia such as the nucleus accumbens (NAcc) and the ventral capsule/ventral striatum (VC/VS) have also been found to have significant antidepressant effects (Anderson et al., 2012; Bewernick et al., 2010, 2012; Malone et al., 2009). As the therapeutic potential of tDCS in psychiatric disorders is further explored, information on how different electrode arrangements determine current density in key brain regions, is essential. This study compared the effects of several DCS montages, with realistic head models reconstructed from MRI head scans, by investigating the brain electric field (E-field) distribution and the average E-field in various brain regions. tDCS montages modelled were those used in recent tDCS depression studies: the F3–supraorbital (F3–SO) montage first used when interest was rekindled in tDCS from 2006 onwards (Boggio et al., 2008; Fregni et al., 2006a, 2006b; Loo et al., 2010; Palm et al., 2011), and modified approaches in which the cathode was moved more laterally to reduce shunting, F3–F8 (Loo et al., 2012), to the right DLPFC, F3–F4 (Brunoni et al., 2011, 2013; Dell'Osso et al., 2012; Ferrucci et al., 2009a, 2009b), or to an extracephalic position to achieve a more widespread pattern of brain activation, F3–extracephalic (F3–EC, brain sites based on the 10–20 EEG system; Martin et al., 2011). The bilateral supraorbital–extracephalic (SO–EC) montage most commonly used in earlier, pre-2000 studies, involving two small anodes at the frontal poles and an extracephalic cathode was also modelled (Arul-Anandam and Loo, 2009; Lippold and Redfearn, 1964; Redfearn et al., 1964). In addition, several hypothetical montages were modelled: supraorbital–occipital (SO–OCC), premised on maximal stimulation of the sgACC and other central and midline subcortical structures; temporal–extracephalic (TMP–EC), prioritising temporal lobe stimulation as neurotrophic changes in this region may have a key role in the pathophysiology of depression (Pittenger and Duman, 2007), and supraorbital–cerebellum (SO– CB), as abnormal cerebellar modulation of the cerebello-thalamocortical pathway has been implicated in the mood and cognitive symptoms associated with several psychiatric disorders, including bipolar disorder and depression (Hoppenbrouwers et al., 2008). The montages were modelled in two subjects — one male and one female — to examine the extent to which inter-individual differences in head anatomy affect variation in electric field with different montages. Finally, a sensitivity analysis was performed to examine the effects of displacing the anodal electrode by ~ 1 cm, to inform on the likely importance of accuracy in electrode placement in clinical applications. Methods Image segmentation and mesh generation Two different high-resolution computational head models were reconstructed from human subjects. One subject was a 35-year-old

333

Asian male whose MRI head scan, labelled “Msub” (short for male subject), was truncated at the level of cervical vertebra 6. The other was a 42-year-old Caucasian female, labelled “Fsub” (female subject, Fig. 1S in Supplementary data): her scan was truncated at the level of the atlas-axis, i.e., cervical vertebrae 1–2. T1-weighted MRI scans of both subjects were obtained from Neuroscience Research Australia. The scans were sagittally-oriented with voxel resolution of 1 mm × 1 mm × 1 mm. The images of Msub were later downsampled to 1.5 mm in every dimension. Head tissue masks were obtained using a combination of automated and manual segmentation softwares. Automated mask generation was performed using BrainSuite, an open-source package from the Laboratory of NeuroImaging at the University of California (Shattuck and Leahy, 2002). Thus, tissue compartments including skin, skull, cerebrospinal fluid (CSF), grey matter (GM) and white matter (WM) were generated from the MRI data. The segmented masks were exported from BrainSuite as grayscale images, and imported into ScanIP (Simpleware Ltd., UK) for manual correction and further processing. The five original masks were hence divided into more compartments: • masks representing eyes, paranasal sinuses, larynx and cervical vertebrae were separated from the skin and skull, as shown in Fig. 1a. In addition, the major foramina of the skull were included in the skull mask, including the superior orbital fissure, optic canal, foramen ovale and foramen magnum; • the skull was divided into the cranium and jaw. The cranium was then subdivided into three layers, with spongy bone tissue as the middle layer, and compact bone tissue as the outermost and innermost layers. The jaw was considered compact. These skull compartments are shown in Fig. 1b; • the brain masks consisted of GM, WM, cerebellum (CB, with brainstem) and the cervical spinal cord (SC), as well as the ventricular system which was later assigned to the CSF mask; • several brain regions of interest (ROIs), considered important in tDCS therapeutic effects, were further segmented from the GM mask as shown in Fig. 1c — anterior cingulate cortices (ACCs), amygdalae, hippocampi, dorsolateral prefrontal cortices (DLPFCs) and orbitofrontal cortices (OFCs). In the head models, fat and muscle were included in the skin compartment, due to the fact that their conductivities are of the same order of magnitude as skin conductivity (Gabriel et al., 1996). Similarly, the venous sinuses and cranial arteries were included in the cerebrospinal fluid compartment. Finally, any remaining blank voxels were manually assigned to the most appropriate neighbouring mask. To examine the effect of an extracephalic clinical electrode montage used in some tDCS studies (Martin et al., 2011; Moliadze et al., 2010), a synthetic upper torso attached to the segmented head was manually painted in ScanIP up to the level above the axilla based on anthropometric measurements (Dreyfuss and Tilley, 1993), as shown in Fig. 1d. A similar approach was used in other studies (Borckardt et al., 2012; Datta et al., 2011, 2012; Mendonca et al., 2011). The + FE Free meshing algorithm in the + FE module of ScanIP (v4.3) was selected to generate the tetrahedral mesh elements for the high-resolution head models, with a compound coarseness of -30. The meshes were then imported into the COMSOL Multiphysics FE solver (v4.2). Tissue conductivities Most compartments of the head models were considered to be electrically homogeneous and isotropic. The electrical conductivity of paranasal sinuses (and larynx) was set to zero. Conductivities of the scalp, compact and spongy bones of the skull, CSF, GM and WM, were assigned mean values from multiple studies (Akhtari et al., 2000, 2002; Baumann et al., 1997; Geddes and Baker, 1967; Gonçalves et al., 2003; Gutierrez et al., 2004; Lai et al., 2005; Oostendorp et al., 2000).

334

S. Bai et al. / NeuroImage 87 (2014) 332–344

brain cerebrospinal fluid

a)

skull

eye

sinuses & pharynx skin spongy bone tissue

vertebrae GM

WM

ACC DLPFC

b)

c) OFC AH compact bone tissue & the jaw

CB SC

d)

Fig. 1. a): Segmentation of the head model “Msub”: skin, eyes, paranasal sinuses (with larynx), skull (including compact bone tissue and spongy bone tissue), vertebrae, CSF and brain. b): Segmentation of skull: compact bone tissue and spongy bone tissue. c): Detailed segmentation of the brain, including defined regions for white matter (WM), grey matter (GM), anterior cingulate cortex, (ACC), dorsolateral prefrontal cortex (DLPFC), orbitofrontal cortex (OFC), amygdala and hippocampus (AH), cerebellum (CB, with brainstem), and cervical spinal cord (SC). d): Frontal view of the model with extended shoulder. To respect the subject's privacy, the eyes of the model are hidden.

All conductivity values are listed in Table 1. The conductivities of the eyes and the synthetic torso were assigned to the scalp conductivity. White matter conductivity anisotropy Modelling studies dedicated to the comparison between isotropic and anisotropic conductivities (Lee et al., 2012; Shahid et al., 2013; Suh et al., 2012), have shown that the presence of WM anisotropy resulted in a significant difference in regional E-fields, especially in the deep brain structures. Hence, the WM anisotropic conductivity was also adopted in this study. DT-MRI was performed only on Msub in 61 gradient directions, with voxel resolution of 2.5 mm × 2.5 mm × 2.5 mm. After being registered to the T1 structural scan in Amira (Visage Imaging GmbH, Germany), diffusion tensor calculation was performed in FSL, an open source software developed by the FMRIB Analysis Group of University of Oxford (Behrens et al., 2003a, 2003b, 2007). Eigenvectors and fractional

Table 1 Tissue conductivities. Compartment

Electrical conductivity (S/m)

Scalp Eyes Sinus CSF GM WM WM (longitudinal) WM (transverse) Vertebrae Synthetic torso Skull (compact bone) Skull (spongy bone)

0.41 0.41 0 1.79 0.31 0.14 0.65 0.065 0.013 0.41 0.006 0.028

S. Bai et al. / NeuroImage 87 (2014) 332–344

Anode

anisotropy (FA) were then calculated, with the latter being widely used to denote the degree of anisotropy: typically greater than 0.45 for WM (Johansen-Berg and Behrens, 2009). The conductivity tensor of WM, σ, was calculated from: ⊤

σ ¼ Sdiagðσ l ; σ t ; σ t ÞS ;

ð1Þ

335

Cathode

F3-SO

where S is the orthogonal matrix of unit eigenvectors obtained from the WM diffusion tensor, and σl and σt are the assigned conductivities longitudinal and transverse to the fibre directions respectively, with σl : σt = 10 : 1 (Nicholson, 1965). σl and σt were calculated using the volume-constraint method (Wolters et al., 2006). This resulted in the values of WM longitudinal and transverse conductivities being 0.65 and 0.065 S/m respectively. Following the diffusion tensor calculation, the calculated conductivity tensors of data points in the DT-MRI scans were then linked to their individual coordinates in the Msub model. This process was performed using MATLAB software (The Mathworks, USA). Only fibre conductivity data having a strong anisotropy signal (FA ≥ 0.45) were exported.

F3-F8

Boundary conditions for volume conductor model

sF3-F8

All head compartments in the tDCS simulations were formulated as passive volume conductors using Laplace's equation: ∇  ð−σ∇φÞ ¼ 0;

ð2Þ

where φ is the electric potential, and σ is the conductivity tensor. For scalp boundaries at the electrodes, two types of boundary conditions were modelled separately:

F3-F4-1

• normal component of inward current density set to Jn for anode, and − Jn for cathode, where J n ¼ area of I electrode, with Is defined as the applied stimulus current fixed at a DC level of 1 mA; • constant voltage set to V for anode, and − V for cathode, with V satisfying ∫ J n dS ¼ I s , where S was the electrode area. s

S

These two types of boundary conditions represented two extreme cases for the electrode. In clinical reality, typical use of a saline-soaked sponge between the skin and electrode pads suggests that neither constant current, or constant voltage conditions are likely to precisely hold at the scalp–electrode interface. Nonetheless, our simulation results suggested that E-field distributions remained generally the same regardless of the choice of electrode boundary condition, even though absolute E-field values were slightly different between the two cases. As a result, we report only results adopting constant current electrode boundaries. Results using fixed voltage electrode boundaries are supplied in the Supplementary data (Fig. 2S). The rest of the boundary conditions were: • ground (zero electric potential) at the lower boundary of the synthetic (extended) torso; • all other external boundaries were assigned as electric insulators (zero normal component of current density); • continuous current density (i.e., flux continuity) across all interior boundaries. Electrode placements Electrodes located on the scalp surface were defined mathematically, which enabled their size, location and orientation to be readily adjusted. A total of eleven tDCS electrode montages were simulated, as shown in Figs. 2 and 3. For most montages, standard 7 cm × 5 cm electrodes were used. For montages with the return electrode placed in an EC position, focal stimulation of the extracephalic region was not required and a 10 cm × 10 cm electrode was used for reasons of comfort. Similarly, a 10 cm × 10 cm return electrode was placed over

F3-F4-2

Fig. 2. tDCS electrode placement — part 1: F3-supraorbital (F3–SO), F3–F8, sF3–F8, F3–F41 and F3–F4-2. The red and blue electrodes represent the anode and cathode, respectively. To respect the subject's privacy, the eyes of the model are hidden.

the occiput for the SO–OCC montage, as the aim was to tailor current pathways to broadly target the medial subcortical structures described above (sgACC, basal ganglia) rather than deliver occipital stimulation. The electrode placements for these montages are described as follows: • F3–SO: both electrodes were 7 cm × 5 cm rectangular pads. The anode was placed over the F3 electrode site on a standard 10–20 EEG cap system, with the long axis of the pad pointing towards the vertex. The cathode was placed above the arcus superciliaris on the right, with the long axis of the pad parallel to the horizontal plane. • F3–F8: both electrodes were 7 cm × 5 cm rectangular pads. The anode was placed at the same location as in F3–SO. The cathode was placed over the F8 EEG cap electrode site, with the lower edge of the pad 4–5 mm below the eye socket, and the long axis perpendicular to the horizontal plane.

336

S. Bai et al. / NeuroImage 87 (2014) 332–344

Anode

Cathode

F3-EC

sF3-EC

TMP-EC

• F3–F4-2: both electrodes were 7 cm × 5 cm rectangular pads, and were placed at the same sites as the F3–F4-1 configuration, but the orientations of both electrodes were perpendicular to those in F3– F4-1. • F3–EC: the 7 cm × 5 cm anode was placed at the same location as in F3–SO, and the cathode was a 10 cm × 10 cm rectangular pad placed on the right shoulder. • sF3–EC: the anode was placed at the same location as that of F3–EC, but the size of the pad was 4 cm × 4 cm. The 10 cm × 10 cm cathode was placed in the same location as the F3–EC configuration. • TMP–EC: the anode was a 7 cm × 5 cm rectangular pad placed between the F3 and T3 EEG positions, with the long axis of the pad parallel to the line connecting F3 and T3. The 10 cm × 10 cm cathode was placed on the right shoulder. • SO–OCC: the anode was a 7 cm × 5 cm rectangular pad, placed above the arcus superciliaris on the left, with the long axis of the pad parallel to the horizontal plane. The cathode was a 10 cm × 10 cm square pad centred over the inion, with the larger electrode size used to diffuse cathodal effects. • SO–CB: the anode was a 7 cm × 5 cm rectangular pad, placed at the same position as the SO–OCC configuration. The cathode was a 10 cm × 5 cm rectangular pad, with the top edge centred at the inion, and the long axis parallel to the horizontal plane, to achieve bilateral cerebellar stimulation. • SO–EC: two circular anodes with a diameter of 0.5 in. (1.27 cm) were placed above the arcus superciliaris on both sides, with the cathode located as in the F3–EC configuration. Data analysis

SO-OCC

SO-CB

SO-EC

Three types of head models were investigated in this study — “Msub-aniso” (referring to the Msub model with WM anisotropy), “Msub” (referring to the Msub model without WM anisotropy) and “Fsub” (without WM anisotropy). The models were solved using the segregated numerical solver in COMSOL (v4.2) on a Windows 64-bit workstation with 24 GB RAM utilising 4 processors. To solve the stationary equations, a direct linear solver was utilised with an absolute error tolerance set to 10−5. It took ~20 min to solve for each simulation, for approximately 1.4 × 106 degrees of freedom. Simulation results were analysed by comparing the difference in brain E-field distribution among the various electrode montages and different head models. The analysis also focused on comparing the average E-field magnitude E in several ROIs in the brain. E was calculated using: Z jEj dV E¼ Z V

;

ð3Þ

dV V

Fig. 3. tDCS electrode placement — part 2: F3-extracephalic (F3–EC), sF3–EC, temporal– extracephalic (TMP–EC), supraorbital–occipital (SO–OCC), supraorbital–cerebellum (SO– CB) and supraorbital–extracephalic (SO–EC). The red and blue electrodes represent the anode and cathode, respectively. To respect the subject's privacy, the eyes of the model are hidden.

• sF3–F8: the anode was placed at the same F3–F8 location, but the size of the anode pad was 4 cm × 4 cm. The cathode was the same as in F3–F8. • F3–F4-1: both electrodes were 7 cm × 5 cm rectangular pads. The anode was placed at the same location as in F3–SO, and the cathode was placed over the F4 electrode site on the 10–20 EEG cap system, with the long axis pointing towards the vertex of the head.

where |E| is the E-field magnitude in the ROI in question, and ∫ V is a volume integral over this region. Note that the denominator is simply the volume of the ROI. The mean of E and its standard error (SE) were then determined across all three head models. The difference in the E-field was considered significant, if the E-field range (mean ± SE) between any two montages did not overlap. In addition, a sensitivity analysis on E-fields in the ROIs to electrode displacement from the original position was conducted on the Msubaniso. Three montages were included in this analysis: F3–F4-1, F3–F8 and F3–EC. For each montage, the anode F3 was perturbed from its original position anteriorly, posteriorly, laterally or medially (note: the antero-posterior displacement was on a sagittal plane, whereas the medio-lateral displacement was on a horizontal plane). The displacement was set to 1 cm, which is the likely margin of error in clinical trials.

S. Bai et al. / NeuroImage 87 (2014) 332–344

The absolute differences in average E-fields from the original placements were calculated. Results Fig. 4 shows the E-field magnitude and direction on the cortical surface of the brain for six selected tDCS electrode montages with Msub-aniso. Fig. 5 shows the E-field profile with the same head model in cross-sectional slices of the brain for the selected tDCS electrode montages. The montages which utilised the F3 anode and a contralateral frontal cephalic cathode (F3–SO, F3–F8, F3–F4-1 as shown in Fig. 4) exhibited higher current density predominately in the frontal lobes of the brain. F3–F8 and F3–F4-1 shared a similar E-field magnitude profile, but the E-field directions were quite distinctive, due to the different positions of the cathodes. As for F3–EC, a more uniform and widely

337

distributed current density was present in the whole brain, especially in the ventral part, owing to current predominantly flowing inferiorly and posteriorly. Compared to the other F3 montages, the E-field magnitude was weaker in the frontal lobe with F3–EC. SO–OCC and TMP–EC had E-field characteristics similar to F3–EC, except that the predominant E-field direction with SO–OCC was anterior–posterior, and the E-field magnitude was large in the left temporal lobe with TMP–EC. The mean and standard error of spatially-averaged E-field magnitude E for the three head models were shown in Fig. 6, with separate data of E in individual head models shown in the Supplementary data (Figs. 3S–5S). In comparison to the F3–SO montage, the modified montages with the anode on F3 but cathode placed at a greater distance from the anode (F3–F8, F3–F4, F3–EC) all resulted in less shunting, i.e. more current entered the brain, with F3–SO significantly different from F3– F8 and F3–EC. The F3–EC montage resulted in the least shunting, with

F3-SO

F3-F8 V/m ≥ 0.25

F3-F4-1 0.15

F3-EC 0.05

SO-OCC

TMP-EC

Fig. 4. E-field magnitude distribution and direction in the whole brain for Msub-aniso model for six selected tDCS montages. The leftmost and two middle columns feature the lateral view from the left, the frontal view and the top view, respectively. The black box specifies the location of the enlarged region shown in the rightmost column. The black arrows represent the direction of E-field vectors, whereas the red arrows indicate the global electrode vector direction — from the anode to the cathode.

338

S. Bai et al. / NeuroImage 87 (2014) 332–344

F3-SO

F3-F8

V/m ≥ 0.25

F3-F4-1

F3-EC

0.15

0.05

SO-OCC

TMP-EC

Fig. 5. E-field brain magnitude distribution in Msub-aniso in cross-sections through the DLPFC (coronal), ACC (sagittal) and OFC (horizontal) with these structures outlined. Dashed lines indicate locations of slice planes.

more current reaching brain areas compared to the other F3 montages. For F3–EC, the E-field in the DLPFCs was reduced compared to other F3 montages, but this montage more strongly modulated the ACCs and cerebellum/brainstem than the other F3 montages. Variation of

electrode size at the F3 site (sF3–F8, sF3–EC) and rotation of electrodes by 90° (F3–F4-2) only made a small difference to modulation patterns. The SO–OCC montage was similar to the F3–EC montage in less shunting and a higher E-field in the ACCs than other frontal montages,

S. Bai et al. / NeuroImage 87 (2014) 332–344

339

V/m 0.1

0

Whole brain

L hemisphere

R hemisphere

CB

L DLPFC

R DLPFC

F3-SO F3-F8 sF3-F8 F3-F4-1 F3-F4-2 F3-EC sF3-EC TMP-EC SO-OCC SO-CB SO-EC

0.1

0

0.1

0

L OFC

R OFC

L ACC

0.1

0

R ACC

L hippocampus

R hippocampus

Fig. 6. Mean and standard error of spatially-averaged E-field magnitude E in various brain regions of interest across the three head models — Msub-aniso, Msub and Fsub.

though the E-field in the DLPFC was reduced. When a smaller cathode was used and centred over the cerebellum (SO–CB), the main effect compared to SO–OCC was an E-field increase in more ventral structures (hippocampus, OFC and CB), but not to a significant level. Compared with more recent frontal montages, the (bilateral) SO–EC resulted in less left DLPFC modulation, but greater OFC and ACC modulation particularly in the left hemisphere. The TMP–EC montage most strongly modulated the marked E-field reduction in frontal structures with TMP–EC. The pattern revealed by the average E-field (E) magnitude with Msub mostly agreed with the anisotropic model. Similarly, the ROI E-field data from Fsub was largely consistent with that from both Msub and Msub-aniso, albeit with minor variations. Namely, Fsub indicated more significant current-shunting in montages with closer inter-electrode scalp distances (E of whole brain, left hemisphere and CB with F3–SO and F3–F4 in Figs. 3S–5S), smaller inter-montage differences in E for left DLPFC, as well as increased E magnitude for right DLPFC with both F3–F8 montages. These discrepancies are likely due to the variation in head geometry, especially in skull size and thickness. Regardless, the ROI E-field magnitude distribution with Fsub was quite similar to that of the corresponding Msub regions (F3–F4-1 in Fig. 7), and there were local hot-spots evident on the cortical surface, but

these were in slightly different locations in each head, due to the variation in cortical foldings. Fig. 8 demonstrates the sensitivity analysis of E-fields (in Msub-aniso) with respect to displacements of the anode in the anterior, posterior, lateral and medial directions. Moving the electrode by 1 cm did not make a substantive difference. Reduction in the inter-electrode distance, such as medial displacement with F3–F4-1 and F3–F8, tended to result in more shunting and less current entering the brain. More significant changes (i.e., close to 10% of the original), were (ROIs): left OFC in F3–F4-1 with antero-posterior displacement, left and right ACCs in F3–F8 with antero-posterior displacement, and left and right ACCs in F3–EC with antero-posterior displacement. Discussion Clinical implications of electrode montages A number of recent tDCS studies have reported positive antidepressant effects in depressed patients (Boggio et al., 2008; Brunoni et al., 2011, 2013; Fregni et al., 2006a, 2006b; Kalu et al., 2012; Loo et al., 2012; Martin et al., 2011; Palm et al., 2011). The rationale guiding the choice of electrode montage in these trials is up-regulation of the left

340

S. Bai et al. / NeuroImage 87 (2014) 332–344

Msub

Fsub

V/m ≥ 0.25

0.15

0.05

Fig. 7. Brain E-field magnitude distribution for the Msub and Fsub models using the F3–F4-1 montage. The rectangles specify the location of the enlarged cortical regions in the bottom row.

DLPFC, with the anode placed on F3 (Boggio et al., 2008; Brunoni et al., 2011, 2013; Dell'Osso et al., 2012; Ferrucci et al., 2009a, 2009b; Fregni et al., 2006a, 2006b; Loo et al., 2010, 2012; Martin et al., 2011; Palm et al., 2011). This is based on the general premise that in depressed patients, the left DLPFC is hypoactive while the right DLPFC is hyperactive (Grimm et al., 2008). The results from the present study showed that all montages with the anode placed at F3 achieved this aim. Findings from the present study also indicate the importance of cathodal placement as it determines the relative current distribution in other brain regions, as well as the intensity of stimulation at the left DLPFC. In comparison to the F3–SO montage where the electrodes are positioned close together, montages with wider spacing between the electrodes result in less shunting and more current entering the brain. However, E-field levels in the DLPFC also decreased as the interelectrode distance increased, in line with Moliadze et al. (2010) which reported a significant decline in motor evoked potentials when they compared left M1-contralateral EC with left M1-contralateral forehead simulations. It is not known whether such a difference is of clinical significance, though a pilot study of the F3–EC montage suggests it may have more rapid antidepressant effects than the F3–F8 montage. This is presumably due to more effective stimulation of other brain regions (Martin et al., 2011), though the relative efficacy of these approaches has yet to be tested in an adequately powered randomised controlled

trial. It is useful to note that use of a smaller electrode at F3 increased the E-field at the left DLPFC. This may be a useful strategy to compensate for loss of intensity associated with moving the cathode further away. The F3–F8 and F3–F4 montages produced similar patterns of the E-field in the brain, resulting in strong stimulation of the left and right DLPFC and clinical trials of these montages have demonstrated antidepressant effects (Brunoni et al., 2011, 2013; Dell'Osso et al., 2012; Ferrucci et al., 2009a; Loo et al., 2012). The role of cathodal stimulation of ventral right frontal regions, particularly achieved with the F3–F8 montage, in antidepressant efficacy is unclear. The F3–F8 montage resulted in a significantly greater E-field in the right OFC compared to the F3–F4 and other montages examined. A recent trial of the F3–F8 montage reported clear antidepressant results, with a 50% response rate after six weeks of tDCS (Loo et al. 2012). However, the independent effects of right frontal cathodal stimulation have not been adequately studied. Boggio et al. (2008) found that the OCC–F8 montage (i.e. cathodal stimulation at F8, used as an active control group) was still able to induce a larger effect than sham tDCS, but results from this trial were inconclusive due to the possible confounding effects of concurrent anodal occipital stimulation and a small sample size. Brain regions other than the DLPFC have also been implicated in depression. These include bilateral sgACC, NAcc, VS/VC, and other central structures, with evidence to suggest that specific aspects of

S. Bai et al. / NeuroImage 87 (2014) 332–344

341

F3-F4-1 F3-F8 F3-EC Post Ant Lat Med

Whole brain

L hemisphere

R hemisphere

CB

L DLPFC

R DLPFC

L OFC

R OFC

L ACC

R ACC

L hippocampus

R hippocampus

Unit: V/m 0

0.04

0.08

0.12

-0.01

0

0.01

Fig. 8. The sensitivity of average E-field magnitude E in brain regions of interest using Msub-aniso, with displacements of F3 electrode with F3–F4-1, F3–F8 and F3–EC. The bar charts on the left represent the average E-field magnitude when F3 is at the original position. The data points on the right represent the absolute differences in average E-fields from the original placement.

depressive symptomatology may be linked to particular brain regions (Anderson et al., 2012). For example, cognitive aspects (pathological guilt, impaired concentration, suicidal ideation) may be linked to ACC, OFC and PFC dysfunction, whereas anhedonia, depressed mood and anxiety may be mediated by the NAcc, sgACC and OFC. Central structures including the hypothalamus, locus coeruleus and dorsal raphe nuclei may be responsible for neurovegetative symptoms such as sleep and appetite disturbances. At present, the relative importance of stimulating such regions remains unclear and it is also unknown whether frontal montages adopted since 2000 are more advantageous in treating depression than the prevailing bilateral SO–EC montage used in earlier decades, which results in strong ACC and OFC activation. The diffuse nature of tDCS may be therapeutically advantageous, given the widespread anatomy of areas involved in depression, as all of the aforementioned areas are stimulated to some extent. This is particularly the case for montages with an extracephalic or posterior electrode, which have been shown to lead to more widespread brain current distribution than montages with a contralateral cephalic cathode.

Theoretically, it is possible for electrode montages to be tailored to an individual patient's depressive symptoms. For example, montages that result in greater stimulation of deep central structures (F3–EC, SO–OCC, SO–CB, SO–EC) may be effective for treating melancholic depression. However, stimulation of the same central structures may increase the risk of a manic episode in bipolar depression (Gálvez et al., 2011). In addition to montages used in clinical practice, several hypothetical montages were also modelled. Though the SO–OCC montage was designed for maximal stimulation of ACC and central structures (thalamus, striatum), our modelling results suggest that ACC modulation is achieved as readily with the F3–EC montage. The mean E-fields in the DLPFC are also similar for the two montages. Given the promising results in an open label depression trial of the F3–EC montage (Martin et al., 2011), a similar antidepressant efficacy is also expected from the SO–OCC montage. High E-fields were found in the cerebellum and hippocampus with the SO–CB and TMP–EC montages respectively, and these montages may be useful where modulation of these brain regions is particularly desired. Both cerebellar and temporal lobe changes have

342

S. Bai et al. / NeuroImage 87 (2014) 332–344

been reported in depression (Fitzgerald et al., 2008; Hoppenbrouwers et al., 2008), and it is possible that neuromodulation targeted at these areas may be beneficial in treating depression. The E-field directions provide a preliminary understanding of the current pathways within the brain, since the current density vector in a conductive medium is in the same direction as the E-field vector. Local E-field vectors generally follow the dominant directionality of the E-field in the entire model, which is from anode to cathode. However, the E-field vectors may not all be in parallel with the main E-field direction, depending on the location of the ROI relative to the electrodes, and on the local electrical properties. For instance, WM fibres provide a preferential pathway for current flow along the fibres when the flow is partially aligned with the fibre orientation, whereas higher E-field magnitudes are generated when current flow is transverse to the fibre orientation (Bai et al., 2013; Lee et al., 2012). Since the modulation of neurons may also be dependent on directionality of the E-field (Basser and Roth, 2000; Creutzfeldt et al., 1962; Molaee-Ardekani et al., 2012), including directionality in analysis may provide more comprehensive information on brain activation, though the clinical implications are as yet unclear. While changes in the electrode size or orientation resulted in differences in the E-field magnitude, the differences were not large. This was perhaps due to a relatively small difference in electrode sizes (35 cm2 vs. 16 cm2), indeed previous studies have shown that large differences in E-field magnitude are only found when there is a considerable change in electrode size (Miranda et al., 2009; Parazzini et al., 2011). Similarly, displacing the electrode by 1 cm around the actual position did not result in a large difference. These results suggest that electrodes may be adjusted slightly to accommodate other priorities, such as skin discomfort and damage. For example, rotation of an electrode to avoid stimulation over a skin lesion such as a mole, may not substantially affect neuronal outcomes. An interesting finding from our results is that current density was not evenly spread over the cerebral cortex, but that local hot spots were evident. This phenomenon has been previously reported in modelling studies (Datta et al., 2009; Salvador et al., 2010) and indicate that these localised areas of intense activation may be of clinical importance for both efficacy and safety. It appears that the location of the hot spot is determined by individual anatomical variations in cerebral folds, thickness of the CSF layer, skull and scalp, and may contribute to interindividual differences in efficacy. Model validation and limitations The novelty of this study is that the anatomically-accurate head models from two different subjects were employed, each with threelayered skull structure and one also with WM anisotropy, in order to investigate several tDCS montages specifically targeted at depression treatment. This was achieved by comparing E-fields in several brain ROIs. Even though some of the above modelling features have appeared in other studies, they have not been combined together in a single study as has been done in this work. In addition, there has been no modelling study dedicated to examining tDCS electrode montages which may be useful in the treatment of depression. As noted by Lee et al. (2012), validation of computer simulations of brain E-field distribution against experimental data remains a challenge. As our head model was only comprised of a limited number of tissues, errors may arise from the exclusion of the dielectric properties of other tissues, even though their electrical conductivities were of the same order as that of the substitute compartment. The conductivities were chosen as an average across different experimental studies, which suggested that they were not measured with the same experimental setup. This, together with the artificial approximation of the torso, may contribute to computational errors. As DTI was not performed on Fsub, Msub also served as a transition model. However, in order to provide a more generalised conclusion for clinical investigation

regarding the effects of these montages, simulations using head models from more subjects with the same electrical properties may be necessary as to further substantiate our results. Nevertheless, the simulation results presented in this study are in line with other modelling studies. Parazzini et al. (2011) reported a median E-field magnitude of 0.33 V/m in the cortex using the left M1– right SO configuration, about five times the average E-field value found here. This is probably due to the variation in head geometry, as well as the adoption of a three-layer skull in our head model, and the difference in the tissue conductivities chosen, which likely resulted in the disparities in the brain E-field magnitude. In addition, Bikson et al. (2010) found a peak cortical E-field value of 0.44 V/m in simulations of a healthy subject with left M1–contralateral forehead tDCS. Datta et al. (2011) simulated the head of a stroke patient, and found a peak value at 0.36 V/m with the left M1–contralateral shoulder montage. Those studies, however, assumed a higher value of skull conductivity than the three-layer skull used in this study. Conclusion With the aid of computational modelling, this study presented a systematic analysis of various tDCS electrode montages that have been used in clinical trials, as well as hypothetical montages specifically for the treatment of depression. The overall profile of E-field magnitude and direction, as well as the average E-field magnitudes in key brain regions were presented. While the clinical significance of these outcomes is not as yet well understood, these results may form a useful reference for researchers seeking to optimise the therapeutic potential of tDCS by manipulation of electrode montage. Sensitivity analysis on the effects of changing the size of the anode, rotating both electrodes and displacing the anode was also conducted on selected montages. This is the first computational study dedicated to this type of analysis. This approach could equally inform the use of tDCS to treat other neuropsychiatric disorders, or the design of montages for neuropsychological experiments. Acknowledgment The authors would like to thank Prof. Caroline Rae and Dr. John Geng from Neuroscience Research Australia for their support in acquiring and processing the structural MRI and DT-MRI data, Dr. Elizabeth Tancred from the University of New South Wales for her expertise in the head anatomy, and Dr. Angelo Alonzo from the Black Dog Institute for his help in determining the accurate position of scalp electrodes. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.neuroimage.2013.11.015. References Akhtari, M., Bryant, H., Mamelak, A., Heller, L., Shih, J., Mandelkern, M., Matlachov, A., Ranken, D., Best, E., Sutherling, W., 2000. Conductivities of three-layer human skull. Brain Topogr. 13, 29–42. Akhtari, M., Bryant, H.C., Mamelak, A.N., Flynn, E.R., Heller, L., Shih, J.J., Mandelkem, M., Matlachov, A., Ranken, D.M., Best, E.D., DiMauro, M.A., Lee, R.R., Sutherling, W.W., 2002. Conductivities of three-layer live human skull. Brain Topogr. 14, 151–167. Anderson, R., Frye, M.A., Abulseoud, O.A., Lee, K.H., McGillivray, J., Berk, M., Tye, S.J., 2012. Deep brain stimulation for treatment-resistant depression: efficacy, safety and mechanisms of action. Neurosci. Biobehav. Rev. 36, 1920–1933. Arul-Anandam, A., Loo, C., 2009. Transcranial direct current stimulation: a new tool for the treatment of depression? J. Affect. Disord. 117, 137–145. Bai, S., Loo, C., Dokos, S., 2013. A review of computational models of transcranial electrical stimulation. Crit. Rev. Biomed. Eng. 41, 21–35. Basser, P., Roth, B., 2000. New currents in electrical stimulation of excitable tissues. Annu. Rev. Biomed. Eng. 2, 377–397. Baumann, S., Wozny, D., Kelly, S., Meno, F., 1997. The electrical conductivity of human cerebrospinal fluid at body temperature. IEEE Trans. Biomed. Eng. 44, 220–223.

S. Bai et al. / NeuroImage 87 (2014) 332–344 Behrens, T., Johansen-Berg, H., Woolrich, M., Smith, S., Wheeler-Kingshott, C., Boulby, P., Barker, G., Sillery, E., Sheehan, K., Ciccarelli, O., Thompson, A., Brady, J., Matthews, P., 2003a. Non-invasive mapping of connections between human thalamus and cortex using diffusion imaging. Nat. Neurosci. 6, 750–757. Behrens, T., Woolrich, M., Jenkinson, M., Johansen-Berg, H., Nunes, R., Clare, S., Matthews, P., Brady, J., Smith, S., 2003b. Characterization and propagation of uncertainty in diffusion-weighted MR imaging. Magn. Reson. Med. 50, 1077–1088. Behrens, T., Berg, H., Jbabdi, S., Rushworth, M., Woolrich, M., 2007. Probabilistic diffusion tractography with multiple fibre orientations: what can we gain? NeuroImage 34, 144–155. Bewernick, B., Hurlemann, R., Matusch, A., Kayser, S., Grubert, C., Hadrysiewicz, B., Axmacher, N., Lemke, M., Cooper-Mahkorn, D., Cohen, M., Brockmann, H., Lenartz, D., Sturm, V., Schlaepfer, T., 2010. Nucleus accumbens deep brain stimulation decreases ratings of depression and anxiety in treatment-resistant depression. Biol. Psychiatry 67, 110–116. Bewernick, B., Kayser, S., Sturm, V., Schlaepfer, T., 2012. Long-term effects of nucleus accumbens deep brain stimulation in treatment-resistant depression: evidence for sustained efficacy. Neuropsychopharmacology 37, 1975–1985. Bikson, M., Datta, A., Rahman, A., Scaturro, J., 2010. Electrode montages for tDCS and weak transcranial electrical stimulation: role of “return” electrode's position and size. Clin. Neurophysiol. 121, 1976–1978. Boggio, P., Rigonatti, S., Ribeiro, R., Myczkowski, M., Nitsche, M., Pascual-Leone, A., Fregni, F., 2008. A randomized, double-blind clinical trial on the efficacy of cortical direct current stimulation for the treatment of major depression. Int. J. Neuropsychopharmacol. 11, 249–254. Borckardt, J., Bikson, M., Frohman, H., Reeves, S., Datta, A., Bansal, V., Madan, A., Barth, K., George, M., 2012. A pilot study of the tolerability and effects of high-definition transcranial direct current stimulation (HD-tDCS) on pain perception. J. Pain 13, 112–120. Brunoni, A., Ferrucci, R., Bortolomasi, M., Vergari, M., Tadini, L., Boggio, P., Giacopuzzi, M., Barbieri, S., Priori, A., 2011. Transcranial direct current stimulation (tDCS) in unipolar vs. bipolar depressive disorder. Prog. Neuropsychopharmacol. Biol. Psychiatry 35, 96–101. Brunoni, A., Valiengo, L., Baccaro, A., Zanão, T., de Oliveira, J., Goulart, A., Boggio, P., Lotufo, P., Benseñor, I., Fregni, F., 2013. The sertraline vs electrical current therapy for treating depression clinical study: results from a factorial, randomized, controlled trial. JAMA Psychiatry 70, 383–391. Carballedo, A., Scheuerecker, J., Meisenzahl, E., Schoepf, V., Bokde, A., Möller, H., Doyle, M., Wiesmann, M., Frodl, T., 2011. Functional connectivity of emotional processing in depression. J. Affect. Disord. 134, 272–279. Creutzfeldt, O., Fromm, G., Kapp, H., 1962. Influence of transcortical DC currents on cortical neuronal activity. Exp. Neurol. 5, 436–452. Datta, A., Elwassif, M., Battaglia, F., Bikson, M., 2008. Transcranial current stimulation focality using disc and ring electrode configurations. J. Neural Eng. 5, 163–174. Datta, A., Bansal, V., Diaz, J., Patel, J., Reato, D., Bikson, M., 2009. Gyri-precise head model of transcranial direct current stimulation: improved spatial focality using a ring electrode versus conventional rectangular pad. Brain Stimul. 2, 201–207. Datta, A., Baker, J.M., Bikson, M., Fridriksson, J., 2011. Individualized model predicts brain current flow during transcranial direct-current stimulation treatment in responsive stroke patient. Brain Stimul. 4, 169–174. Datta, A., Dmochowski, J., Guleyupoglu, B., Bikson, M., Fregni, F., 2012. Cranial electrotherapy stimulation and transcranial pulsed current stimulation: a computer based highresolution modeling study. NeuroImage 65, 280–287. Dell'Osso, B., Zanoni, S., Ferrucci, R., Vergari, M., Castellano, F., D'Urso, N., Dobrea, C., Benatti, B., Arici, C., Priori, A., Altamura, A., 2012. Transcranial direct current stimulation for the outpatient treatment of poor-responder depressed patients. Eur. Psychiatry 27, 513–517. Dreyfuss, H., Tilley, A., 1993. Measure of Man and Woman: Human Factors in Design. John Wiley & Sons, NY. Ferrucci, R., Bortolomasi, M., Brunoni, A., Vergari, M., Tadini, L., Giacopuzzi, M., Priori, A., 2009a. Comparative benefits of transcranial direct current stimulation (tDCS) treatment in patients with mild/moderate vs. severe depression. Clin. Neuropsychiatry 6, 246–251. Ferrucci, R., Bortolomasi, M., Vergari, M., Tadini, L., Salvoro, B., Giacopuzzi, M., Barbieri, S., Priori, A., 2009b. Transcranial direct current stimulation in severe, drug-resistant major depression. J. Affect. Disord. 118, 215–219. Fitzgerald, P., Laird, A., Maller, J., Daskalakis, Z., 2008. A meta-analytic study of changes in brain activation in depression. Hum. Brain Mapp. 29, 683–695. Fox, M., Buckner, R., White, M., Greicius, M., Pascual-Leone, A., 2012. Efficacy of transcranial magnetic stimulation targets for depression is related to intrinsic functional connectivity with the subgenual cingulate. Biol. Psychiatry 72, 595–603. Fregni, F., Boggio, P., Nitsche, M., Marcolin, M., Rigonatti, S., Pascual-Leone, A., 2006a. Treatment of major depression with transcranial direct current stimulation. Bipolar Disord. 8, 203–204. Fregni, F., Boggio, P., Nitsche, M., Rigonatti, S., Pascual-Leone, A., 2006b. Cognitive effects of repeated sessions of transcranial direct current stimulation in patients with depression. Depress. Anxiety 23, 482–484. Gabriel, C., Gabriel, S., Corthout, E., 1996. The dielectric properties of biological tissues: I. Literature survey. Phys. Med. Biol. 41, 2231–2249. Gálvez, V., Alonzo, A., Martin, D., Mitchell, P.B., Sachdev, P., Loo, C.K., 2011. Hypomania induction in a patient with bipolar II disorder by transcranial direct current stimulation (tDCS). J. ECT 27, 256–258. Geddes, L., Baker, L., 1967. The specific resistance of biological material — a compendium of data for the biomedical engineer and physiologist. Med. Biol. Eng. 5, 271–293.

343

Gonçalves, S., de Munck, J., Verbunt, J., Bijma, F., Heethaar, R., Lopes da Silva, F., 2003. In vivo measurement of the brain and skull resistivities using an EIT-based method and realistic models for the head. IEEE Trans. Biomed. Eng. 50, 754–767. Greicius, M., Flores, B., Menon, V., Glover, G., Solvason, H., Kenna, H., Reiss, A., Schatzberg, A., 2007. Resting-state functional connectivity in major depression: abnormally increased contributions from subgenual cingulate cortex and thalamus. Biol. Psychiatry 62, 429–437. Grimm, S., Beck, J., Schuepbach, D., Hell, D., Boesiger, P., Bermpohl, F., Niehaus, L., Boeker, H., Northoff, G., 2008. Imbalance between left and right dorsolateral prefrontal cortex in major depression is linked to negative emotional judgment: an fMRI study in severe major depressive disorder. Biol. Psychiatry 63, 369–376. Gutierrez, D., Nehorai, A., Muravchik, C., 2004. Estimating brain conductivities and dipole source signals with EEG arrays. IEEE Trans. Biomed. Eng. 51, 2113–2122. Hoppenbrouwers, S., Schutter, D., Fitzgerald, P., Chen, R., Daskalakis, Z., 2008. The role of the cerebellum in the pathophysiology and treatment of neuropsychiatric disorders: a review. Brain Res. Rev. 59, 185–200. Johansen-Berg, H., Behrens, T., 2009. Diffusion MRI: From Quantitative Measurement to In-vivo Neuroanatomy. Academic Press, London. Kalu, U., Sexton, C., Loo, C., Ebmeier, K., 2012. Transcranial direct current stimulation in the treatment of major depression: a meta-analysis. Psychol. Med. 1, 1–10. Lai, Y., Van Drongelen, W., Ding, L., Hecox, K., Towle, V., Frim, D., He, B., 2005. Estimation of in vivo human brain-to-skull conductivity ratio from simultaneous extraand intra-cranial electrical potential recordings. Clin. Neurophysiol. 116, 456–465. Lee, W., Deng, Z., Kim, T., Laine, A., Lisanby, S., Peterchev, A., 2012. Regional electric field induced by electroconvulsive therapy in a realistic finite element head model: influence of white matter anisotropic conductivity. NeuroImage 59, 2110–2123. Lippold, O., Redfearn, J., 1964. Mental changes resulting from the passage of small direct currents through the human brain. Br. J. Psychiatry 110, 768–772. Loo, C., Sachdev, P., Martin, D., Pigot, M., Alonzo, A., Malhi, G., Lagopoulos, J., Mitchell, P., 2010. A double-blind, sham-controlled trial of transcranial direct current stimulation for the treatment of depression. Int. J. Neuropsychopharmacol. 13, 61–69. Loo, C., Alonzo, A., Martin, D., Mitchell, P., Galvez, V., Sachdev, P., 2012. Transcranial direct current stimulation for depression: 3-week, randomised, sham-controlled trial. Br. J. Psychiatry 200, 52–59. Lozano, A., Giacobbe, P., Hamani, C., Rizvi, S., Kennedy, S., Kolivakis, T., Debonnel, G., Sadikot, A., Lam, R., Howard, A., Ilcewicz-Klimek, M., Honey, C., Mayberg, H., 2012. A multicenter pilot study of subcallosal cingulate area deep brain stimulation for treatment-resistant depression. J. Neurosurg. 116, 315–322. Malone, D., Dougherty, D., Rezai, A., Carpenter, L., Friehs, G., Eskandar, E., Rauch, S., Rasmussen, S., Machado, A., Kubu, C., Tyrka, A., Price, L., Stypulkowski, P., Giftakis, J., Rise, M., Malloy, P., Salloway, S., Greenberg, B., 2009. Deep brain stimulation of the ventral capsule/ventral striatum for treatment-resistant depression. Biol. Psychiatry 65, 267–275. Martin, D., Alonzo, A., Mitchell, P., Sachdev, P., Gálvez, V., Loo, C., 2011. Frontoextracephalic transcranial direct current stimulation as a treatment for major depression: an open-label pilot study. J. Affect. Disord. 134, 459–463. Mayberg, H., 1997. Limbic–cortical dysregulation: a proposed model of depression. J. Neuropsychiatry Clin. Neurosci. 9, 471–481. Mayberg, H., 2003. Modulating dysfunctional limbic–cortical circuits in depression: towards development of brain-based algorithms for diagnosis and optimised treatment. Br. Med. Bull. 65, 193–207. Mayberg, H., 2007. Defining the neural circuitry of depression: toward a new nosology with therapeutic implications. Biol. Psychiatry 61, 729–730. Mayberg, H., Lozano, A., Voon, V., McNeely, H., Seminowicz, D., Hamani, C., Schwalb, J., Kennedy, S., 2005. Deep brain stimulation for treatment-resistant depression. Neuron 45, 651–660. Mendonca, M., Santana, M., Baptista, A., Datta, A., Bikson, M., Fregni, F., Araujo, C., 2011. Transcranial dc stimulation in fibromyalgia: optimized cortical target supported by high-resolution computational models. J. Pain 12, 610–617. Miranda, P., Lomarev, M., Hallett, M., 2006. Modeling the current distribution during transcranial direct current stimulation. Clin. Neurophysiol. 117, 1623–1629. Miranda, P., Faria, P., Hallett, M., 2009. What does the ratio of injected current to electrode area tell us about current density in the brain during tDCS? Clin. Neurophysiol. 120, 1183–1187. Molaee-Ardekani, B., Márquez-Ruiz, J., Merlet, I., Leal-Campanario, R., Gruart, A., SánchezCampusano, R., Birot, G., Ruffini, G., Delgado-García, J., Wendling, F., 2012. Effects of transcranial direct current stimulation (tDCS) on cortical activity: a computational modeling study. Brain Stimul. http://dx.doi.org/10.1016/j.brs.2011.12.006. Moliadze, V., Antal, A., Paulus, W., 2010. Electrode-distance dependent after-effects of transcranial direct and random noise stimulation with extracephalic reference electrodes. Clin. Neurophysiol. 121, 2165–2171. Nicholson, P., 1965. Specific impedance of cerebral white matter. Exp. Neurol. 13, 386–401. Nitsche, M., Paulus, W., 2000. Excitability changes induced in the human motor cortex by weak transcranial direct current stimulation. J. Physiol. 527, 633–639. Nitsche, M., Paulus, W., 2001. Sustained excitability elevations induced by transcranial DC motor cortex stimulation in humans. Neurology 57, 1899–1901. Oostendorp, T., Delbeke, J., Stegeman, D., 2000. The conductivity of the human skull: results of in vivo and in vitro measurements. IEEE Trans. Biomed. Eng. 47, 1487–1492. Palm, U., Schiller, C., Fintescu, Z., Obermeier, M., Keeser, D., Reisinger, E., Pogarell, O., Nitsche, M., Möller, H., Padberg, F., 2011. Transcranial direct current stimulation in treatment resistant depression: a randomized double-blind, placebo-controlled study. Brain Stimul. http://dx.doi.org/10.1016/j.brs.2011.08.005.

344

S. Bai et al. / NeuroImage 87 (2014) 332–344

Parazzini, M., Fiocchi, S., Rossi, E., Paglialonga, A., Ravazzani, P., 2011. Transcranial direct current stimulation: estimation of the electric field and of the current density in an anatomical human head model. IEEE Trans. Biomed. Eng. 58, 1773–1780. Pittenger, C., Duman, R.S., 2007. Stress, depression, and neuroplasticity: a convergence of mechanisms. Neuropsychopharmacology 33, 88–109. Redfearn, J., Lippold, O., Costain, R., 1964. Preliminary account of the clinical effects of polarizing the brain in certain psychiatric disorders. Br. J. Psychiatry 110, 773–785. Salvador, R., Mekonnen, A., Ruffini, G., Miranda, P., 2010. Modeling the electric field induced in a high resolution realistic head model during transcranial current stimulation. Conf Proc IEEE Eng Med Biol Soc, pp. 2073–2076. Shahid, S., Wen, P., Ahfock, T., 2013. Numerical investigation of white matter anisotropic conductivity in defining current distribution under tDCS. Comput. Methods Programs Biomed. 109, 48–64.

Shattuck, D., Leahy, R., 2002. BrainSuite: an automated cortical surface identification tool. Med. Image Anal. 6, 129–142. Suh, H.S., Lee, W.H., Kim, T.S., 2012. Influence of anisotropic conductivity in the skull and white matter on transcranial direct current stimulation via an anatomically realistic finite element head model. Phys. Med. Biol. 57, 6961–6980. Wagner, T., Fregni, F., Fecteau, S., Grodzinsky, A., Zahn, M., Pascual-Leone, A., 2007. Transcranial direct current stimulation: a computer-based human model study. NeuroImage 35, 1113–1124. Weaver, L., Williams, R., Rush, S., 1976. Current density in bilateral and unilateral ECT. Biol. Psychiatry 11, 303–312. Wolters, C., Anwander, A., Tricoche, X., Weinstein, D., Koch, M., MacLeod, R., 2006. Influence of tissue conductivity anisotropy on EEG/MEG field and return current computation in a realistic head model: a simulation and visualization study using highresolution finite element modeling. NeuroImage 30, 813–826.