Effects of Transcranial Direct Current Stimulation on Attentional Bias to Methamphetamine Cues and Its Association With EEG-Derived Functional Brain Network Topology

Abstract

Background Although transcranial direct-current stimulation (tDCS) has shown to potentially mitigate drug craving and attentional bias to drug-related stimuli, individual differences in such modulatory effects of tDCS are less understood. In this study, we aimed to investigate a source of the inter-subject variability in the tDCS effects that can be useful for tDCS-based treatments of individuals with methamphetamine use disorder (IMUD). Methods Forty two IMUD (all male) were randomly assigned to receive a single-session of either sham or real bilateral tDCS (anodal right/cathodal left) over the dorsolateral prefrontal cortex. The tDCS effect on methamphetamine craving and biased attention to drug stimuli were investigated by quantifying EEG-derived P3 (a measure of initial attentional bias) and late positive potential (LPP; a measure of sustanined motivated attention), elicited by these stimuli. To assess the association of changes in P3 and LPP with brain connectivity network (BCN) topology, correlation between topology metrics, specifically those related to the effficientcy of information processing, and the tDCS effect was investigated. Results The P3 amplitude significantly decreased following the tDCS session, whereas in the sham group the amplitudes increased. The changes in P3 amplitudes were significantly correlated with communication efficiency measured by BCN topology metrics (r=-0.47, p=0.03; r=-0.49, p=0.02). There was no significant change in LPP amplitude due to the tDCS application. Conclusions These findings validate that tDCS mitigates initial attentional bias, but not the sustained motivated attention, to methamphetamine stimuli. Importantly, however, results also show that the individual differences in the effects of tDCS may be underpinned by communication efficiency of the BCN topology, and therefore, these BCN topology metrics may have the potential to robustly predict the effectiveness of tDCS-based interventions on MA craving and attentional bias to MA stimuli among IMUD.

A scoping review of electroencephalographic (EEG) markers for tracking neurophysiological changes and predicting outcomes in substance use disorder treatment

“…One study examined the effects of 20 sessions with 5-min iTBS (theta, 3-pulse 50 Hz bursts, left dlPFC) over 4 weeks and found no changes in methamphetamine vs. neutral cue ERPs but indicated decreases in frontal-central beta (Chen et al, 2021 ). A second study with one session of 20-min tDCS (2 mA, left dlPFC) reported stimulation-related decreased drug vs. neutral P3 (at post-stimulation only) that was correlated with decreased resting-state graph-theoretical (i.e., network) efficiency in multiple areas including bilateral medial temporal, left superior parietal and angular gyrus, and right frontal and supplementary motor regions (Khajehpour et al, 2022 ). A spectral study examining the effects of 20 sessions with 4-min iTBS (theta, 3-pulse 50 Hz, left dlPFC) over 10 days found decreased left prefrontal delta that was linked to treatment-related increases in working memory performance (Zhang et al, 2021a ).…”

Fusion Graph Representation of EEG for Emotion Recognition

“…Papers [ 2 , 3 , 4 , 5 , 6 ] utilize the topology structure to construct the graph. Jang et al [ 2 ] used the physical distance between EEG electrodes to obtain an adjacency matrix containing intra-band and inter-band connections, successfully represented EEG data as a graphed signal, and applied it to video recognition based on EEG.…”

“…Introducing graph theory to brain networks, Chen et al [ 4 ] used a minimum spanning tree to generate a topology graph according to link strength. Paper [ 5 ] considered brain topology metrics based on graph theory, which determined local and global efficiency. Duan et al [ 6 ] considered the number of steps required to get from one node to another in a brain network, and defined the average of the shortest paths between any two nodes in the network as the topological connection.…”

“…Researchers developed various graph neural network (GNN) models to overcome this limitation by projecting EEG electrodes onto the graph nodes, updating graph edge weights, and applying the graph to EEG-based tasks. Studies on topology connection [ 2 , 3 , 4 , 5 , 6 ], functional connection [ 7 , 8 , 9 , 10 , 11 , 12 , 13 ], and effective connection [ 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 ] have proved their superiority in extracting helpful information on channel relationships for EEG emotion recognition. Topology connection is to measure the adjacency of channels in physical distance.…”

Challenges and future trends in wearable closed-loop neuromodulation to efficiently treat methamphetamine addiction

“…As EEG records neural activity on a millisecond timescale, the neural signals triggered by stimuli (visual, audio, etc.) show specific wave forms, namely the eventrelated potential (ERP) (111,112).…”