Introduction

The application of therapeutic electrical currents for neuromodulation has attracted growing interest among neurotherapists. Transcranial direct current stimulation (tDCS) is a non-invasive technique used to facilitate or inhibit neural activity in both pathological and healthy conditions. When determining the stimulation dose, clinicians must consider the targeted cortical location, inter-individual variability (e.g., tissue conductivity, head geometry, medical history) and the desired therapeutic outcomes. Two key biophysical parameters in tDCS are the induced electric field strength and the resulting current distribution. Electric field strength (E) denotes the intensity of the field generated by the tDCS electrode montage and is typically measured in volts per meter (V/m). Within tDCS, E is governed by the potential difference between the anode (positive) and cathode (negative) electrodes, and the field extends from the anode to the cathode. The magnitude of the electric field strength is influenced by various factors, including electrode size, placement, current intensity, and tissue conductivity. Current distribution describes the pathway of charge flow through cerebral and extracerebral tissues during stimulation; current travels from the anode to the cathode, creating a direct-current circuit that depolarizes or hyperpolarizes neurons in the underlying brain regions. The distribution is non-uniform and depends on head-tissue geometry and conductivity, as well as electrode size, shape, and placement. In vivo tDCS studies often employ computational modeling to predict intracranial electric field strength and current distribution. These models incorporate the individual’s head anatomy, electrode placement, and the electrical properties of scalp, skull, cerebrospinal fluid, and brain tissue to estimate the field distribution. Researchers use these models to optimize tDCS protocols and ensure that the targeted regions receive the desired stimulation while minimizing off-target effects. A sound understanding of electric field strength and current distribution is critical for both safety and efficacy in tDCS applications, as different levels and patterns of stimulation can produce distinct neuromodulatory effects 1.

Several proprietary and open-source software packages are available to characterize and analyze electric field strength and current distribution, including COMSOL Multiphysics, the Realistic Volumetric Approach to Simulate Transcranial Electric Stimulation (ROAST) 2, Simulation of Non-Invasive Brain Stimulation (SimNIBS) 3, SPHEARES 4, and SCIRun 5. Among these platforms, ROAST and SimNIBS are the most widely cited—**225 and 800 citations, respectively, up to April 2022—**for modeling electric fields and current distributions 6. Mohigul Nasimova et al. evaluated ROAST for clinical validation and reported that it is more robust than SimNIBS 6. Both SimNIBS and ROAST use MRI-derived three-dimensional head models to simulate tDCS in individual subjects 7. SimNIBS additionally supports simulations of transcranial alternating current stimulation (tACS), in which sinusoidal currents modulate cortical neurons. Between 2019 and 2023, two major versions of SimNIBS (3.0 and 4.0) and 32 incremental updates were released, adding new utilities, enhanced operating-system compatibility, and security patches (Figure 1).

With respect to features and workflow, SimNIBS can be compared with other freeware packages such as ROAST, COMET, BONSAI, and SPHEARES. SimNIBS, ROAST, and COMET provide automatic segmentation of T1- and T2-weighted MRI, standard conductivity assignment, electrode placement, three-dimensional mesh generation, and finite-element solvers. BONSAI supports only electrode placement for visualization and cannot perform MRI-based E-field analysis. SPHEARES lacks automatic T1/T2 segmentation, resulting in non-patient-specific E-field visualizations. Table 1 summarizes the minimum system requirements for the free tDCS simulation software.

In this review, we focused on studies that employed SimNIBS across diverse clinical investigations. These investigations analyzed multiple parameters to model and compare therapeutic outcomes of tDCS in neurological disorders. We report stimulation dose, target site, neurological indication, and other relevant factors incorporated into in vivo modeling.

Methods

Results

We selected 100 research articles from Google Scholar, filtering results published between 2019 and 2025 by year and keyword relevance, because including all 711 initially retrieved papers would have rendered the review excessively long. Table 2 details the various neurorehabilitation and neuromodulation applications of transcranial direct-current stimulation (tDCS) that incorporated the SimNIBS modeling software in single or multiple analyses. Specifically, Table 2 summarizes the SimNIBS version employed, clinical application area, patient characteristics, sample size, targeted cortical region, electrode montage, electric-field (E-field) metrics, and reported therapeutic outcomes. The reviewed studies clustered into eight domains: Cognitive Enhancement and Neuromodulation; Neurological and Psychiatric Disorders; Smoking Cessation and Substance Use; Physical Performance and Rehabilitation; Electrode and Electric-Field Analysis; Memory and Attention; Neuroimaging and Brain Structure; and Age-Related Research (Figure 2). Across the 100 articles, 87 distinct sub-domains were identified. Figure 3 depicts the annual distribution of publications from 2019 to 2023.

Andrés Molero-Chamizo et al. administered five anodal tDCS sessions to the motor cortex of three stroke survivors and observed improvements in pain control and reductions in upper-limb spasticity, albeit with marked inter-individual variability 8. Paulo J. C. Suen et al. correlated simulated E-field strength with behavioral change in 16 patients with major depressive disorder receiving tDCS to the dorsolateral prefrontal cortex (DLPFC) and anterior cingulate cortex (ACC); higher E-field intensity was inversely associated with negative affect in both regions and with depression scores in the left ACC, suggesting a mechanistic link that merits further study 9. A cross-diagnostic investigation reported diminished prefrontal E-field strength during tDCS in individuals with schizophrenia and, to a lesser extent, major depressive disorder, whereas patients with bipolar disorder showed no significant difference from healthy controls 10. Karin Prillinger et al. explored the feasibility, tolerability, and neural effects of tDCS in adolescents with autism-spectrum disorder, focusing on social and emotional functioning, and highlighted the modality’s therapeutic potential 11.

One study investigated the potential of transcranial direct current stimulation (tDCS) to enhance motor learning in children with perinatal stroke (PS) by simulating tDCS-induced electric fields (EFs) with various electrode montages. The authors reported montage-dependent differences in EF strength and its relationship to underlying anatomy in children with arterial ischemic stroke (AIS) or periventricular infarction (PVI) compared with typically developing controls (TDC), thereby emphasizing the importance of individualized tDCS planning for future clinical trials 12. In efforts to optimize multichannel tDCS, Videira et al. showed that electrode spacing—particularly the anode-to-cathode distance—exerts the greatest influence on EF distribution, whereas the use of more than three cathodes produced no additional change in EF magnitude or direction. These findings inform efficient electrode placement, especially during concurrent tDCS-electroencephalography (EEG) recordings 13. Mizutani-Tiebel et al. assessed individual, MRI-derived electric fields (e-fields) during standard bifrontal tDCS in 74 participants with major depressive disorder (MDD), schizophrenia (SCZ), or healthy status. They identified significant differences in e-field strength between clinical and non-clinical groups, highlighting the need for individualized dosing in patient populations 14. Two separate randomized clinical trials targeted nicotine craving and relapse. tDCS applied to the dorsolateral prefrontal cortex (DLPFC) significantly reduced nicotine dependence and the likelihood of future smoking 15,16. In both studies, tDCS simulations performed in SimNIBS estimated EF strength and focality within the DLPFC under distinct stimulation parameters. Although SimNIBS provides fully automated head-tissue segmentation, its developers caution that accuracy may be compromised in the presence of structural brain pathology. Huang et al. addressed this limitation by assigning lesion-specific conductivity values and comparing the resulting EFs across different modeling platforms, including SimNIBS and ROAST 17. Because glutamate is the principal excitatory neurotransmitter, tDCS-induced currents may modulate its release and regulation, thereby altering synaptic plasticity. Mezger et al. combined EF simulations with proton magnetic resonance spectroscopy to quantify glutamate concentrations and resting-state functional connectivity during tDCS 18. Additional findings were presented at the 2022 Neuroergonomics and NYC Neuromodulation Conference, but detailed results have yet to be published.

Athena Stein et al. conducted a comparative analysis of electric field simulations in children who were cognitively normal or had mild or severe traumatic brain injury (TBI) 19. They found no significant differences in electric field strength across groups, indicating that current delivery was comparable. In psychiatric disorders such as anorexia nervosa (AN) and obsessive–compulsive disorder (OCD), transcranial direct current stimulation (tDCS) has demonstrated therapeutic efficacy, reducing symptoms over multiple sessions. Studies 20,21 simulated electric field strength with SimNIBS on a standard head model and obtained 0.368 V/m for AN and 1 V/m for OCD. In a cue-reactivity paradigm for methamphetamine use disorder (MUD), an average electric field of 0.35 V/m was associated with modulation during cue exposure 22,23. Using the SimNIBS standard head models ‘ernie’ and MNI152, investigators estimated electric field strength to assess physiological and cognitive effects of tDCS applied to premotor areas and the dorsolateral prefrontal cortex (DLPFC) 24,25. In healthy young adults, no significant improvements in working memory were observed 26,27; however, Kevin A. Caulfield et al. suggested that reducing cortical electric-field variance may enhance outcomes 28. The DLPFC, located on the lateral aspect of the prefrontal cortex, is critical for executive functions such as planning, reasoning, and problem-solving. Five independent studies targeting cognitive enhancement with tDCS reported improved performance 29,30,31,32,33. These investigations used MNI coordinates and the international 10–20 EEG system to position the electrodes and simulated electric fields on standard head models for conceptualization.

In the realm of mood disorders, M. A. Bertocci et al. reported that applying cathodal tDCS over the ventrolateral prefrontal cortex (vlPFC) is an effective therapeutic intervention for bipolar disorder 34. Similarly, Zhenhong He et al. have shown that anodal tDCS over the right vlPFC reduces subjective emotional intensity and physiological arousal elicited by negative experiences 35.

Within psychotic disorders, amelioration of symptoms in schizophrenia and schizophrenia-spectrum disorders has been achieved by administering tDCS to the frontopolar cortex 36 and parietal lobe 37.

For impulse-control and behavioural disorders, Ilse Verveer et al. demonstrated the ability of dACC-targeted HD-tDCS to modulate neurophysiological indices of impulsivity 38. Carmen S. Sergiou and colleagues reported increased frontal connectivity within the alpha- and beta-frequency bands following HD-tDCS 39. In a separate study, the same group found that repeated HD-tDCS sessions applied to the ventromedial prefrontal cortex (vmPFC) decreased aggression 40.

Regarding cognitive control, cognitive performance, and physical endurance, tDCS applied to the dorsolateral prefrontal cortex (DLPFC) has proven beneficial 41,42,43.

The parietal lobe contributes substantially to multiple components of memory processing. It participates in the encoding, consolidation, and retrieval of episodic memories—representations of specific events or experiences—and supports spatial memory, enabling individuals to navigate and retain the spatial configuration of their surroundings. Electrode montages that specifically stimulate the parietal cortex have been shown to improve associative memory, attentional control, and lexical retrieval 44,45,46. Transcranial direct current stimulation (tDCS) may also augment verbal fluency and diminish mind-wandering in patients with attention-deficit/hyperactivity disorder (ADHD) 47,48. The sustained neuromodulatory effects of tDCS observed in cognitive rehabilitation indicate a promising therapeutic avenue both for post-COVID-19 cognitive impairment and for preserving cognitive health in older adults without dementia 49,50,51.

Computational modeling of transcranial direct current stimulation (tDCS) entails simulating and analyzing the spatial distribution and physiological effects of electric currents delivered to the brain. This process typically employs computer algorithms and mathematical models to predict the electric field distribution, current density, and overall impact on neural activity within the targeted cortical regions. The overarching goal is to elucidate the potential physiological and cognitive outcomes of tDCS under various stimulation parameters, electrode montages, and individual anatomical variations. Consequently, computational modeling is indispensable for optimizing tDCS protocols and for clarifying the mechanisms underlying its neuromodulatory effects.

Numerous scholars have contributed to this field, frequently using SimNIBS to develop or test stimulation protocols. Carys Evans et al. simulated three T1-weighted MRI head models with ROAST and SimNIBS, highlighting the importance of simulation software in characterizing inward and outward current flow 52. Andrés Molero-Chamizo et al. used SimNIBS and the COMETS platform to simulate twenty typical electrode configurations, reporting no substantial differences in the resulting electric fields between the two tools 53. Utkarsh Pancholi et al. examined the influence of electrode size and shape, current intensity, and electrode gel on the simulated electric field and focality, observing a notable reduction in both metrics as electrode size increased 54,55. Similarly, M. A. Callejón-Leblic et al. targeted diverse brain regions and documented significant changes in electric-field strength and distribution, particularly in peak tangential and normal components 56.

In a methodological comparison, Gaurav V. Bhalerao and colleagues evaluated segmentation and head-modeling pipelines—SimNIBS Freesurfer-FSL (mri2mesh: SNF), SimNIBS headreco (CAT12: SNC), SimNIBS headreco (SPM: SNS), ROAST (RST), and ScanIP Abaqus (ABQ)—across thirty-two subjects and quantified the relative differences in mean electric fields within predefined regions of interest 57. Likewise, Kevin A. Caulfield and collaborators optimized electrode positioning (traditional vs. innovative), electrode size (larger vs. smaller), and inter-electrode distance (greater vs. lesser) to enhance electric-field magnitude and focality in the targeted cortex 58. Two additional investigations focused on improving electric-field characteristics through montage optimization, specifically by adjusting electrode positions and their spatial relationships 59,60.

Discussion

It is important to note that the computational analysis of therapeutic outcomes from both in vivo and in vitro transcranial direct current stimulation (tDCS) studies is increasingly important. In-silico platforms such as SimNIBS, ROAST, and COMET have streamlined these analyses by offering detailed models of cranial tissues and their electrical properties when exposed to direct current. SimNIBS, in particular, is more widely used than other simulation platforms. It can be used for predictive modelling with standard head templates in clinical trials to anticipate therapeutic efficacy at early stages 11,20,22,26,31,41,43,45,50. The increasing application of electric-field (E-field) analysis in SimNIBS entails systematic variation of stimulation parameters and the implementation of optimization algorithms. Such analyses forecast current distribution and focality within the target region. For instance, optimization algorithms have been applied to predict differential tDCS effects in cognitively normal (CN) individuals and in patients with mild cognitive impairment (MCI) or Alzheimer’s disease (AD) 1. Likewise, modifications in electrode size, shape, or conductive medium alter induced current density and clinical outcomes 54. SimNIBS also quantifies the normal and tangential components of the induced E-field during tDCS 56. Current-flow modelling allows estimation and optimization of inward and outward current orientations, thereby localizing zones of depolarization and hyperpolarization 52,53,106. E-field patterns vary across individuals, a phenomenon that can be systematically investigated with computational modelling 23,59. These models also enable comparison between SimNIBS and other in-silico platforms 57.

In this study, we identified the locations of the stimulating electrodes on standard or MRI-derived head models using the 10–20/10–10 EEG placement system and MNI-based target coordinates (Table 2, Column 7). Column 6 of Table 2 lists the corresponding brain/head regions where the electrodes were positioned or where the electric field was predicted. Column 5 reports the total number of participants (n = 3856) who took part in clinical studies or simulations utilizing the SimNIBS platform, underscoring the platform’s versatility and ongoing development. In addition to common neurological and psychiatric disorders, SimNIBS has been employed in a diverse range of applications, including Parkinson’s disease 63, neurotransmitter quantification 64, episodic migraine prophylaxis 65, assessment of skin–cerebellar distance and morphometric posterior fossa parameters 67, chronic low-back pain 68, sleepiness and vigilance 69, temporal perception 70, early dexterity skills 75, Stroop error analysis 78, and conscious error correction 82. Figure 4 illustrates typical bipolar electrode configurations: anode and cathode placement on a three-dimensional head model, their interface with grey matter, electric-field heat-mapping (indicating red as maximum and blue as minimum E-field strength), as well as HD-tDCS placement on a 3D head model and grey matter interface, demonstrating the placement of stimulating electrodes in patients.

The electric field intensity is a very important parameter for tDCS, measured in V/m, and represents the strength of the electrical force acting on charged particles (ions) within brain tissue. This metric is critical because it modulates neuronal activity. The magnitude of the field determines the degree to which neuronal membranes become polarized or depolarized, thereby altering neural excitability. However, the relationship between field intensity and physiological outcome is complex; electrode size, placement, and inter-individual variability also contribute. In the present review we assessed the reported electric field intensity in each of the 100 articles examined and observed wide variability across studies and objectives. For example, A. Molero-Chamizo et al. used electric field modelling to reach a peak intensity of 0.36 V/m in a standard head model, aiming to modulate spasticity while accommodating inter-individual variability 8. Participants with schizophrenia or major depressive disorder displayed significantly lower e-field strength at the 99.5th percentile than healthy controls 10. Several studies evaluated the reliability of electric-field estimates by comparing software packages; SimNIBS showed greater reliability than COMET 53,57. Variations in tDCS current intensity produce differential effects on cortical blood flow, consequently modifying motor-cortex excitability 90. Table 2, column 9, summarises the reported electric field intensities, expressed as peak or mean values in V/m or mV/mm.

The primary aim of this review is to demonstrate the influence of SimNIBS and its applicability in devising effective treatment regimens through predefined parameter selection and modeling. No significant limitations have been identified regarding the functionality or effectiveness of SimNIBS. The outcomes of the 100 reviewed clinical investigations are primarily affected by the sample size, the number of electrodes, the inter-electrode distance, inter-individual variability, electrode size and shape, electrode thickness, stimulation site, patient age, sex, and the conductivities of cranial tissues. Molero-Chamizo et al. reported on three stroke patients enrolled in a 9- to 15-month study that evaluated the therapeutic effect—specifically pain relief—after several sessions of tDCS 8. The small sample size is additionally concerning because pain was recorded only immediately before and after each of the five interventions, with no follow-up assessments.

In a separate study, Andreia S. Videira et al. investigated how inter-electrode distance and electrode number influence the induced electric field, illustrating E-field fluctuations in a standard brain/head model to quantify variability 13. Including additional subjects could introduce confounding factors, as the objective was solely to examine E-field variation attributable to electrode distance and number. Parisa Banaei et al. used standard, large, rectangular tDCS electrodes to simulate an optimal electric field (>0.20–0.25 V/m) 31. The goal of this simulation study was to replicate the desired electric field, necessitating deliberate selection of electrode dimensions, shape, and placement. However, the scarcity of electrical modeling tools for injured or lesioned brains and the ongoing challenges in T1 anatomical image reconstruction constrain such investigations. Eva Mezger et al. therefore recommended further research and development of reconstruction algorithms and field models for damaged brain structures in neuropsychiatric and neurological disorders, including stroke rehabilitation 36.

Molero-Chamizo et al. employed SimNIBS and COMETS to test 20 tDCS electrode configurations with a finite-element approach, comparing the results and assessing inter-individual variability 53. Either a subject-specific realistic model is sufficient to evaluate inter-individual variability. Automated tissue segmentation of T1- and T2-weighted MRI data is a core component of SimNIBS. However, normal tissue contrast may be disrupted in pathological conditions such as stroke, tumors, traumatic brain injury, or neurodegenerative diseases 111. Such contrast alterations can lead to misclassification of CSF, edema, or necrotic tissue, and failure to model cavities or calcifications accurately. Complex pathological anatomies, including enlarged ventricles, cortical atrophy, or tissue displacement, are difficult for SimNIBS to mesh and simulate accurately, potentially resulting in erroneous electric field estimates 112. SimNIBS relies on literature-derived average values and therefore assumes static, isotropic tissue conductivities that may not reflect the altered, anisotropic, and dynamic properties of lesioned tissues under pathological conditions 113. Moreover, SimNIBS has not yet undergone extensive validation in clinical populations; limited evidence exists comparing simulated with measured electric fields or evaluating the impact of simulation-guided therapies in diseased brains 114.

Conclusion

In summary, SimNIBS has garnered considerable interest among neurophysicists, biomedical engineers, and other research professionals owing to its versatile functionality and robust computational capabilities. It provides accurate predictions of the spatial distribution and focality of the electric field prior to the administration of transcranial direct current stimulation (tDCS), thereby preventing unintended stimulation and enhancing therapeutic outcomes. The peak or mean electric-field intensity generated during tDCS typically ranges from 0.5 to 2 V m⁻¹; however, these values are modulated by several variables, including applied current strength, electrode size and montage, and inter-individual anatomical variability. Such influences can be estimated with in-silico platforms such as SimNIBS. We identified 100 highly relevant research articles published between 2019 and 2023; an additional 611 publications may also contain information pertinent to the usability of this current-flow modelling platform.

The absence of standardized simulation methodologies continues to hamper tDCS modelling, leading to heterogeneous findings across studies. Future investigations should harmonize electrode montages, current intensities, and modelling parameters to improve reproducibility and enable cross-study comparisons. The development of best-practice guidelines, particularly for SimNIBS, is expected to enhance consistency and clinical applicability. While current head models accurately approximate current flow in healthy individuals, models for pathological conditions—especially stroke, epilepsy, and traumatic brain injury—require further refinement. Subsequent research should incorporate disease-specific anatomical alterations, such as lesions, atrophy, and abnormal tissue conductivities, to optimise clinical stimulation protocols and treatment planning. Finally, rigorous validation of computational simulations against experimental and clinical data remains imperative. Integrating patient-specific computational models with empirical measures, including neuroimaging (e.g., fMRI, BrainSuite) and electrophysiology (e.g., EEG), is anticipated to improve the predictive accuracy of tDCS models and facilitate their translation to clinical practice.

Abbreviations

A: Anode; ACC: Anterior cingulate cortex; ADHD: Attention deficit hyperactivity disorder; AIS: Arterial ischemic stroke; AN: Anorexia nervosa; ASD: Autism spectrum disorder; AUD: Alcohol use disorder; C: Cathode; CAI: Chronic Ankle Instability; CAT: Computational anatomical toolbox; COMETS: Computation of Electric field due to transcranial current Stimulation; COMT: Catechol-O-Methyltransferase; dACC: Dorsal anterior cingulate cortex; DC: Direct current; DSM: Diagnostic and Statistical Manual of Mental Disorders; EEG: Electroencephalography; EF: Electric field; E-field: Electric field; GABA: Gamma-aminobutyric acid; Glx: Glutamate and glutamine levels; HC: Healthy control; HCP: Human connectome project; ICD: International Classification of Diseases; IFG: Inferior frontal gyrus; LBP: Lower back pain; LTA: Lateral temporoparietal area; mA: Milliampere; MDD: Major depressive disorder; MNI: Montreal neuroimaging initiative; mTBI: Mild traumatic brain injury; MUD: Methamphetamine use disorder; n: Number; OCD: Obsessive-compulsive disorder; PMI: Primary motor cortex; PS: Perinatal stroke; PVI: Periventricular infarction; RPE: Rating of perceived exertion; SCZ: Schizophrenia; SD: Standard deviation; SimNIBS: Simulation of non-invasive brain stimulation; SMN: Sensorimotor Network; svTBI: Severe traumatic brain injury; tACS: Transcranial alternating current stimulation; TBI: Traumatic brain injuries; tDCS: Transcranial direct current stimulation; TEP: TMS-evoked potentials; TMS: Transcranial magnetic stimulation; TP: Temporoparietal; V/m: volts per meter; Ver: Version; VL: Vastus lateralis; vlPFC: Ventrolateral prefrontal cortex; vPFC: Ventral prefrontal cortex; YBOCS: Yale-brown obsessive compulsive score

Acknowledgments

None