Autism Spectrum Disorder (ASD) encompasses a diverse range of neurodevelopmental conditions characterized by challenges in social communication, along with the presence of stereotyped, repetitive behaviors, and restricted interests. These symptoms, stemming from atypical nervous system development, emerged in early childhood, yet the pathogenesis remained elusive (1). Currently, interventions for children with ASD primarily consist of conventional rehabilitation therapies aimed at altering psychological states and rectifying abnormal behaviors through psychological and behavioral means to facilitate treatment outcomes. Nonetheless, targeted therapies for ASD are still under active investigation.

The emergence of non-invasive stimulation techniques in autism research has propelled transcranial direct current stimulation (tDCS) to the forefront. tDCS delivers a steady stream of electrical current to the cerebral cortex, modulating cortical excitability—it is both safe and non-invasive (2). tDCS influences neuronal excitability and plasticity by administering electrical currents and offers benefits such as mild stimulation intensity, affordability, and a reduced risk of adverse effects. Its application spans cognitive impairment, depression, and Alzheimer’s disease treatments and rehabilitation research (3, 4) and showed promise for addressing cognitive deficits associated with neurological conditions (5). Importantly, tDCS does not require prolonged stillness from children, making it particularly suitable for children with ASD who may exhibit low functioning and hyperactivity. The modality also includes sham stimulation settings, accommodating the demands of clinical research. Cumulative evidence suggests that applying tDCS to the frontal cortex can markedly influence brain activities, including attention (6), learning (6–8), memory (7–9), and vigilance (10). These profound effects position tDCS as a potential therapeutic modality for ASD, although the efficacy of tDCS in children with ASD requires further exploration.

Resting-state Electroencephalogram (EEG) can reveal intrinsically connected functional networks, unobscured by the demands of cognitive tasks. While connectivity studies in EEG have traditionally relied on measurements between electrode pairs, this approach potentially suffers from volume conduction issues (11) and imprecise anatomical location estimates, assuming a two-dimensional construct rather than a true three-dimensional framework. Conversely, three-dimensional analysis utilizing EEG may furnish more intricate and accurate insights.

EEG source localization aims to identify sources of neural activity in the brain through potential differences between EEG electrodes on the scalp. An important development was the introduction of anatomical constraints on the head to facilitate the problem of EEG source location (12, 13). Further developments introduced physiological constraints of cortical sources to facilitate the solution of distributed source imaging (14). This prior constraint greatly improves the solvability and precision of brain source location. The distributed source model is based on thousands of orientationally fixed dipole sources that are regularly distributed in a three-dimensional (3D) brain space or cortical surface. The activation of the source is described as a current density distribution. One advantage of the distributed model is that there is no need to make a prior assumption about the number of active dipoles under a particular scalp voltage diagram. The minimum norm algorithm was proposed in 1994 to estimate the source current distribution as an alternative to single or multiple dipole models. Since the minimum norm tends to mislocate deeper sources, the weighted minimum norm method can compensate for this with a weighted matrix. Low resolution electromagnetic Tomography (LORETA) aims to find the smoothest minimum norm solution to the EEG activity distribution, standardized LORETA (sLORETA) standardizes the current density estimate given by the minimum norm by using the covariance of the resolution matrix and is able to achieve zero error localization. We can examine EEG data through the lens of current source density measurements, made possible by standardized low-resolution electromagnetic tomography (sLORETA). This methodology synergized the high temporal resolution of EEG with the spatial localization of cerebral electrical activities (15). Correlation analysis revealed augmented short-range connectivity between regions implicated in the mirror neuron system and social perception networks (16). A previous magnetoencephalography investigation into ASD employed source localization to assess resting-state networks (RSN), including the default mode and salience networks (17). The findings associated reductions in gamma-band connectivity within the DMN and between the DMN and salience networks with ASD severity, particularly concerning social communication and interaction capacities. Historical focus has been scant on the sensorimotor network (SMN), which underlies core ASD symptoms due to atypical sensorimotor and perceptual processes (18–24). Moreover, studies regarding the functional brain networks sustaining attentional faculties are rare. The dorsal attention network (DAN) orchestrates goal-driven top-down control processes (25) and deficits in the connectivity within this network may underpin the attentional challenges characteristic of ASD (26, 27).

Hence, based on the above evidence, we conducted a shame-controlled trial to investigate the efficacy of tDCS in children with ASD. This study hypothesized that (1) significant differences of current source density in delta, theta, alpha, and beta rhythm in the DMN, SMN, and DAN would be observed between children with ASD and typically developing (TD) children. (2) active tDCS stimulation would increase current source density in the above described resting state networks and rhythms in the brain of children with ASD. (3) active tDCS stimulation would enhance functional connectivity (lagged phase synchronization) of resting-state networks in the brain of children with ASD.

In this study, we evaluated the variations in current source density across the delta, theta, alpha, and beta bands within the brains of children diagnosed with ASD in comparison to typically developing children, as well as the changes in ASD children before and after undergoing tDCS stimulation. Utilizing lagged phase synchronization via sLORETA, we’ve examined functional connectivity alterations within three brain resting states before and after tDCS. Our findings indicated a notably lower current source density in the neural architecture of children in the ASD group when compared to the TD group. Post-tDCS regulation, the active stimulation group exhibited a substantial enhancement in brain current source density, which wasn’t observed in the sham stimulation group. Notably, ASD children’s brain activation improved significantly following tDCS intervention, with an increased number and distribution of voxels showing significant differences.

There is a well-documented array of evidence suggesting that individuals with ASD possess atypical brain structures and functioning (37–39), and non-invasive brain stimulation methods such as tDCS have demonstrated a considerable impact on modulating both brain activity and connective function (40, 41). Active tDCS stimulation, in contrast with sham, has been shown to reduce resting-state functional connections (42). Correspondingly, our study recorded a noteworthy rise in lagged phase synchronization within the DMN, SMN, and DAN in children with ASD post tDCS stimulation.

Existing research indicates that anomalous intrinsic functional connectivity of brain networks is a central aspect of social impairments in ASD (43). Moreover, the DMN—which plays a crucial role in processing self and other-related information and is implicated in recollecting episodic and autobiographical memories (44), as well as understanding others’ mental states (45)—has emerged as a pivotal system contributing to social deficits in ASD (46, 47). In line with this, our study discerned enhanced connectivity in the DMN between the medial frontal gyrus and bilateral superior temporal gyres in children with ASD post tDCS, which may significantly remodel internal information processing, potentially easing social impairments in ASD (48).

As a neural sensor, the SMN is primarily engaged in the handling and transmission of sensory and motor information. The interconnections between the different areas within this network ensure precise coordination of sensory and motor function. Diminished connectivity within the somatosensory and motor networks corroborates the atypical multisensorial and motor integration seen in ASD patients, likely contributing to core symptoms (49, 50). Our findings suggest that tDCS stimulation may elevate responsiveness to sensory input in individuals with ASD, thereby enhancing multisensory integration and consequently improving social capacities (51–53).

The DAN is recognized as a bilateral network that sustains attentional stability and contributes to the varying aspects of intelligence development (25, 54). Given the plethora of sensory signals the brain continually receives; DAN enables prioritization of attention to the foremost inputs at any given moment (55). Aberrations in DAN functional connectivity can manifest as impaired attention and hindered top-down processing in ASD. Following stimulation, increased connectivity within key DAN regions in our study might bolster attentional control and top-down goal-directed processing in individuals with ASD (25).

Thus, our research demonstrates that tDCS stimulation is safe and reliable, and may amplify brain activity and augment functional connectivity in individuals with ASD, potentially ameliorating core symptoms. And our study shows that tDCS can be applied to clinical treatment for a long time, and has far-reaching significance to improve the status quo of autistic children. Nevertheless, the employment of tDCS as a therapeutic approach for ASD remains in its nascent stages, with our study facing certain limitations. Firstly, our research depended on a relatively modest sample size with broader samples required in future inquiries. Additionally, evaluations were only conducted at two points—baseline and post-treatment. Subsequent research should incorporate multiple assessments, including follow-ups, to capture critical trajectories of tDCS’ effects. Thirdly, our study solely focused on the left DLPFC as a target area for tDCS, despite emerging studies that suggest various potential targets each potentially addressing specific ASD symptoms. Future investigations must explore these alternative sites. Lastly, while sLORETA serves as our means to measure brain activity and delineate regions, it is not without spatial limitations.

In conclusion, this study employed sLORETA to elucidate differences in brain activity between children with ASD and their TD counterparts, demonstrating that tDCS intervention in the DLPFC can enhance brain connectivity and evidencing the positive neuromodulatory effects of tDCS. Conferring the possibility that tDCS may emerge as a promising therapeutic avenue for children with ASD, its clinical application could become increasingly widespread, adding valuable practicality in clinical settings.

The raw data supporting the conclusions of this article will be made available by the corresponding author on reasonable request.

The studies involving humans were approved by the Ethics Committee of the Affiliated Hospital of Hebei University. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants’ legal guardians/next of kin.

XL: Writing – review & editing, Funding acquisition, Conceptualization. JK: Writing – original draft, Supervision. YL: Writing – review & editing, Formal analysis, Data curation. SL: Writing – review & editing, Software, Investigation, Data curation. PH: Writing – review & editing, Supervision, Methodology.