Investigating neuromodulatory effect of transauricular vagus nerve stimulation on resting-state electroencephalography.

Purpose: The purpose of this study was to investigate the neuromodulatory effects of transauricular vagus nerve stimulation (taVNS) and determine optimal taVNS duration to induce the meaningful neuromodulatroty effects using resting-state electroencephalography (EEG). Method: Fifteen participants participated in this study and taVNS was applied to the cymba conchae for a duration of 40 min. Resting-state EEG was measured before and during taVNS application. EEG power spectral density (PSD) and brain network indices (clustering coefficient and path length) were calculated across five frequency bands (delta, theta, alpha, beta and gamma), respectively, to assess the neuromodulatory effect of taVNS. Moreover, we divided the whole brain region into the five regions of interest (frontal, central, left temporal, right temporal, and occipital) to confirm the neuromodulation effect on each specific brain region. Result: Our results demonstrated a significant increase in EEG frequency powers across all five frequency bands during taVNS. Furthermore, significant changes in network indices were observed in the theta and gamma bands compared to the pre-taVNS measurements. These effects were particularly pronounced after approximately 10 min of stimulation, with a more dominant impact observed after approximately 20-30 min of taVNS application. Conclusion: The findings of this study indicate that taVNS can effectively modulate the brain activity, thereby exerting significant effects on brain characteristics. Moreover, taVNS duration of approximately 20-30 min was considered appropriate for inducing a stable and efficient neuromodulatory effects. Consequently, these findings have the potential to contribute to research aimed at enhancing cognitive and motor functions through the modulation of EEG using taVNS. Supplementary Information: The online version contains supplementary material available at 10.1007/s13534-024-00361-8.

Differences in Functional Connectivity between Patients with Depression with and without Nonsuicidal Self-injury.

Objective: : Nonsuicidal self-injury (NSSI), which involves deliberate harm to body tissues without suicidal intent, represents an escalating clinical concern. We used electroencephalography (EEG) to investigate the differences in functional connectivity (FC) patterns in patients with depression with and without a history of NSSI. Methods: : Seventy-seven patients with mood disorders experiencing major depressive episodes were categorized into NSSI (Group A; n = 31) and non-NSSI (Group B; n = 46) groups on the basis of their NSSI history. EEG data were collected and FC was analyzed using coherence (Coh), imaginary coherence (iCoh), and phase-locking value (PLV) metrics. Network indices based on graph theory were calculated. Demographic and clinical characteristics and scale scores were compared between groups A and B. Results: : While the two groups showed no significant differences in demographic characteristics such as age and diagnosis, the Beck Depression Inventory and Suicidal Ideation Questionnaire (SIQ) scores were higher in Group A. Binary logistic regression analyses revealed associations of NSSI with sex and the SIQ score. We examined the connectivity of 1,326 pairs of signals across six frequency bands, yielding 7,956 signal pairs. The two groups showed no significant differences in the Coh, iCoh, corrected PLV, or network indices but showed significant differences in all the frequency bands when an uncorrected t test was used. Conclusion: : In this study, FC differences in depression with and without NSSI were not observed. Further well-controlled research is expected to clarify neurobiological underpinnings and guide future interventions.

Risk of data leakage in estimating the diagnostic performance of a deep-learning-based computer-aided system for psychiatric disorders.

Deep-learning approaches with data augmentation have been widely used when developing neuroimaging-based computer-aided diagnosis (CAD) systems. To prevent the inflated diagnostic performance caused by data leakage, a correct cross-validation (CV) method should be employed, but this has been still overlooked in recent deep-learning-based CAD studies. The goal of this study was to investigate the impact of correct and incorrect CV methods on the diagnostic performance of deep-learning-based CAD systems after data augmentation. To this end, resting-state electroencephalogram (EEG) data recorded from post-traumatic stress disorder patients and healthy controls were augmented using a cropping method with different window sizes, respectively. Four different CV approaches were used to estimate the diagnostic performance of the CAD system, i.e., subject-wise CV (sCV), overlapped sCV (oSCV), trial-wise CV (tCV), and overlapped tCV (otCV). Diagnostic performances were evaluated using two deep-learning models based on convolutional neural network. Data augmentation can increase the performance with all CVs, but inflated diagnostic performances were observed when using incorrect CVs (tCV and otCV) due to data leakage. Therefore, the correct CV (sCV and osCV) should be used to develop a deep-learning-based CAD system. We expect that our investigation can provide deep-insight for researchers who plan to develop neuroimaging-based CAD systems for psychiatric disorders using deep-learning algorithms with data augmentation.

Relationship between Loudness Dependence of Auditory-evoked Potentials and Suicidality.

Objective: : Serotonin concentration is associated with suicide in patients with major depressive disorder. Loudness dependence of auditory-evoked potentials (LDAEPs), a representative neurophysiological indicator, is related to serotonin activity. Therefore, this study aimed to investigate the relationship between LDAEPs and suicidal ideation, suicide attempts, and the severity of depression. Methods: : We evaluated the scalp N1, P2, and N1/P2 LDAEPs along with standardized low-resolution brain electromagnetic tomography (sLORETA)-localized N1, P2, and N1/P2 LDAEPs of 221 patients with major depressive disorder. The demographic and clinical data of the patients, including data on suicidal ideation and previous suicide attempts, were obtained from clinical interviews and medical records. The severity of depression was assessed using the Beck Depression Inventory and Hamilton Depression Rating Scale, whereas suicidal ideation was evaluated using the Beck Scale for Suicidal Ideation (BSS). Results: : The total BSS score was associated with low N1/P2 LDAEP (p = 0.045), whereas P2 sLORETA-LDAEP was associated with lower previous suicide attempts (p = 0.043). In addition, suicide attempt was correlated with an elevated P2 left sLORETA-LDAEP (coefficient = 4.638, p = 0.038). Conclusion: : These findings suggest that suicidal ideation is associated with decreased LDAEP, whereas suicide attempt is associated with increased LDAEP.

Electroencephalography Network Indices as Biomarkers of Upper Limb Impairment in Chronic Stroke.

Alteration of electroencephalography (EEG) signals during task-specific movement of the impaired limb has been reported as a potential biomarker for the severity of motor impairment and for the prediction of motor recovery in individuals with stroke. When implementing EEG experiments, detailed paradigms and well-organized experiment protocols are required to obtain robust and interpretable results. In this protocol, we illustrate a task-specific paradigm with upper limb movement and methods and techniques needed for the acquisition and analysis of EEG data. The paradigm consists of 1 min of rest followed by 10 trials comprising alternating 5 s and 3 s of resting and task (hand extension)-states, respectively, over 4 sessions. EEG signals were acquired using 32 Ag/AgCl scalp electrodes at a sampling rate of 1,000 Hz. Event-related spectral perturbation analysis associated with limb movement and functional network analyses at the global level in the low-beta (12-20 Hz) frequency band were performed. Representative results showed an alteration of the functional network of low-beta EEG frequency bands during movement of the impaired upper limb, and the altered functional network was associated with the degree of motor impairment in chronic stroke patients. The results demonstrate the feasibility of the experimental paradigm in EEG measurements during upper limb movement in individuals with stroke. Further research using this paradigm is needed to determine the potential value of EEG signals as biomarkers of motor impairment and recovery.

Neuromodulatory feasibility of a current limiter-based tDCS device: a resting-state electroencephalography study.

Recently, we introduced a current limiter-based novel transcranial direct-current stimulation (tDCS) device that does not generate significant tDCS-induced electrical artifacts, thereby facilitating simultaneous electroencephalography (EEG) measurement during tDCS application. In this study, we investigated the neuromodulatory effect of the tDCS device using resting-state EEG data measured during tDCS application in terms of EEG power spectral densities (PSD) and brain network indices (clustering coefficient and path length). Resting-state EEG data were recorded from 10 healthy subjects during both eyes-open (EO) and eyes-closed (EC) states for each of five different conditions (baseline, sham, post-sham, tDCS, and post-tDCS). In the tDCS condition, tDCS was applied for 12 min with a current intensity of 1.5 mA, whereas tDCS was applied only for the first 30 s in the sham condition. EEG PSD and brain network indices were computed for the alpha frequency band most closely associated with resting-state EEG. Both alpha PSD and network indices were found to significantly increase during and after tDCS application compared to those of the baseline condition in the EO state, but not in the EC state owing to the ceiling effect. Our results demonstrate the neuromodulatory effect of the tDCS device that does not generate significant tDCS-induced electrical artifacts, thereby allowing simultaneous measurement of electrical brain activity. We expect our novel tDCS device to be practically useful in exploring the impact of tDCS on neuromodulation more precisely using ongoing EEG data simultaneously measured during tDCS application.

Toward practical machine-learning-based diagnosis for drug-naïve women with major depressive disorder using EEG channel reduction approach.

BACKGROUND: A machine-learning-based computer-aided diagnosis (CAD) system can complement the traditional diagnostic error for major depressive disorder (MDD) using trait-like neurophysiological biomarkers. Previous studies have shown that the CAD system has the potential to differentiate between female MDD patients and healthy controls. The aim of this study was to develop a practically useful resting-state electroencephalography (EEG)-based CAD system to assist in the diagnosis of drug-naïve female MDD patients by considering both the drug and gender effects. In addition, the feasibility of the practical use of the resting-state EEG-based CAD system was evaluated using a channel reduction approach. METHODS: Eyes-closed, resting-state EEG data were recorded from 49 drug-naïve female MDD patients and 49 sex-matched healthy controls. Six different EEG feature sets were extracted: power spectrum densities (PSDs), phase-locking values (PLVs), and network indices for both sensor- and source-level, and four different EEG channel montages (62, 30, 19, and 10-channels) were designed to investigate the channel reduction effects in terms of classification performance. RESULTS: The classification performances for each feature set were evaluated using a support vector machine with leave-one-out cross-validation. The optimum classification performance was achieved when using sensor-level PLVs (accuracy: 83.67 % and area under curve: 0.92). Moreover, the classification performance was maintained until the number of EEG channels was reduced to 19 (over 80 % accuracy). CONCLUSION: We demonstrated the promising potential of sensor-level PLVs as diagnostic features when developing a resting-state EEG-based CAD system for the diagnosis of drug-naïve female MDD patients and verified the feasibility of the practical use of the developed resting-state EEG-based CAD system using the channel reduction approach.

ETSNet: A deep neural network for EEG-based temporal-spatial pattern recognition in psychiatric disorder and emotional distress classification.

The use of EEG for evaluating and diagnosing neurological abnormalities related to psychiatric diseases and identifying human emotions has been improved by deep learning advancements. This research aims to categorize individuals with schizophrenia (SZ), their biological relatives (REL), and healthy controls (HC) using resting EEG brain source signal data defined by regions of interest (ROIs). The proposed solution is a deep neural network for the cortical source signals of the ROIs, incorporating a Squeeze-and-Excitation Block and multiple CNNs designed for eyes-open and closed resting states. The model, called EEG Temporal Spatial Network (ETSNet), has two variants: ETSNets and ETSNetf. Two evaluations were conducted to show the effectiveness of the proposed model. The average accuracy for the classification of SZ, REL, and HC using EEG resting data was 99.57% (ETSNetf), and the average accuracy for the classification of eyes-open (EO) and eyes-closed (EC) resting states was 93.15% (ETSNets). An ablation study was also conducted using two public datasets for intellectual and developmental disorders and emotional states, showing improved classification accuracy compared to advanced EEG classification algorithms when using ETSNets.

Altered Functional Networks of Alpha and Low-Beta Bands During Upper Limb Movement and Association with Motor Impairment in Chronic Stroke.

Background: Impaired movement after stroke is closely associated with altered brain functions, and thus the investigation on neural substrates of patients with stroke can pave a way for not only understanding the underlying mechanisms of neuropathological traits, but also providing an innovative solution for stroke rehabilitation. The objective of this study was to precisely investigate altered brain functions in terms of power spectral and brain network analyses. Methods: Altered brain function was investigated by using electroencephalography (EEG) measured while 34 patients with chronic stroke performed movement tasks with the affected and unaffected hands. The relationships between functional brain network indices and Fugl-Meyer Assessment (FMA) scores were also investigated. Results: A stronger low-beta event-related desynchronization was found in the contralesional hemisphere for both affected and unaffected movement tasks compared with that of the ipsilesional hemisphere. More efficient whole-brain networks (increased strength and clustering coefficient, and prolonged path length) in the low-beta frequency band were revealed when moving the unaffected hand compared with when moving the affected hand. In addition, the brain network indices of the contralesional hemisphere indicated higher efficiency and cost-effectiveness than those of the ipsilesional hemisphere in both the alpha and low-beta frequency bands. Moreover, the alpha network indices (strength, clustering coefficient, path length, and small-worldness) were significantly correlated with the FMA scores. Conclusions: Efficient functional brain network indices are associated with better motor outcomes in patients with stroke and could be useful biomarkers to monitor stroke recovery during rehabilitation.

Impaired functional cortical networks in the theta frequency band of patients with post-traumatic stress disorder during auditory-cognitive processing.

Impaired cognitive function related to intrusive memories of traumatic experiences is the most noticeable characteristic of post-traumatic stress disorder (PTSD); nevertheless, the brain mechanism involved in the cognitive processing is still elusive. To improve the understanding of the neuropathology in PTSD patients, we investigated functional cortical networks that are based on graph theory, by using electroencephalogram (EEG). EEG signals, elicited by an auditory oddball paradigm, were recorded from 53 PTSD patients and 39 healthy controls (HCs). Source signals in 68 regions of interests were estimated using EEG data for each subject using minimum-norm estimation. Then, using source signals of each subject, time-frequency analysis was conducted, and a functional connectivity matrix was constructed using the imaginary part of coherence, which was used to evaluate three global-level (strength, clustering coefficient, and path length) and two nodal-level (strength and clustering coefficients) network indices in four frequency bands (theta, alpha, low-beta, and high-beta). The relationships between the network indices and symptoms were evaluated using Pearson's correlation. Compared with HCs, PTSD patients showed significantly reduced spectral powers around P300 periods and significantly altered network indices (diminished strength and clustering coefficient, and prolonged path length) in theta frequency band. In addition, the nodal strengths and nodal clustering coefficients in theta band of PTSD patients were significantly reduced, compared with those of HCs, and the reduced nodal clustering coefficients in parieto-temporo-occipital regions had negative correlations with the symptom scores (Impact of Event Scale-Revises, Beck Depression Inventory, and Beck Anxiety Inventory). The characterization of this disrupted pattern improves the understanding of the neuropathophysiology underlying the impaired cognitive function in PTSD patients.

Enhanced Performance by Interpretable Low-Frequency Electroencephalogram Oscillations in the Machine Learning-Based Diagnosis of Post-traumatic Stress Disorder.

Electroencephalography (EEG)-based diagnosis of psychiatric diseases using machine-learning approaches has made possible the objective diagnosis of various psychiatric diseases. The objective of this study was to improve the performance of a resting-state EEG-based computer-aided diagnosis (CAD) system to diagnose post-traumatic stress disorder (PTSD), by optimizing the frequency bands used to extract EEG features. We used eyes-closed resting-state EEG data recorded from 77 PTSD patients and 58 healthy controls (HC). Source-level power spectrum densities (PSDs) of the resting-state EEG data were extracted from 6 frequency bands (delta, theta, alpha, low-beta, high-beta, and gamma), and the PSD features of each frequency band and their combinations were independently used to discriminate PTSD and HC. The classification performance was evaluated using support vector machine with leave-one-out cross validation. The PSD features extracted from slower-frequency bands (delta and theta) showed significantly higher classification performance than those of relatively higher-frequency bands. The best classification performance was achieved when using delta PSD features (86.61%), which was significantly higher than that reported in a recent study by about 13%. The PSD features selected to obtain better classification performances could be explained from a neurophysiological point of view, demonstrating the promising potential to develop a clinically reliable EEG-based CAD system for PTSD diagnosis.

Inflated prediction accuracy of neuropsychiatric biomarkers caused by data leakage in feature selection.

In recent years, machine learning techniques have been frequently applied to uncovering neuropsychiatric biomarkers with the aim of accurately diagnosing neuropsychiatric diseases and predicting treatment prognosis. However, many studies did not perform cross validation (CV) when using machine learning techniques, or others performed CV in an incorrect manner, leading to significantly biased results due to overfitting problem. The aim of this study is to investigate the impact of CV on the prediction performance of neuropsychiatric biomarkers, in particular, for feature selection performed with high-dimensional features. To this end, we evaluated prediction performances using both simulation data and actual electroencephalography (EEG) data. The overall prediction accuracies of the feature selection method performed outside of CV were considerably higher than those of the feature selection method performed within CV for both the simulation and actual EEG data. The differences between the prediction accuracies of the two feature selection approaches can be thought of as the amount of overfitting due to selection bias. Our results indicate the importance of correctly using CV to avoid biased results of prediction performance of neuropsychiatric biomarkers.

Resting-State Functional Connectivity in Mathematical Expertise.

To what extent are different levels of expertise reflected in the functional connectivity of the brain? We addressed this question by using resting-state functional magnetic resonance imaging (fMRI) in mathematicians versus non-mathematicians. To this end, we investigated how the two groups of participants differ in the correlation of their spontaneous blood oxygen level-dependent fluctuations across the whole brain regions during resting state. Moreover, by using the classification algorithm in machine learning, we investigated whether the resting-state fMRI networks between mathematicians and non-mathematicians were distinguished depending on features of functional connectivity. We showed diverging involvement of the frontal-thalamic-temporal connections for mathematicians and the medial-frontal areas to precuneus and the lateral orbital gyrus to thalamus connections for non-mathematicians. Moreover, mathematicians who had higher scores in mathematical knowledge showed a weaker connection strength between the left and right caudate nucleus, demonstrating the connections' characteristics related to mathematical expertise. Separate functional networks between the two groups were validated with a maximum classification accuracy of 91.19% using the distinct resting-state fMRI-based functional connectivity features. We suggest the advantageous role of preconfigured resting-state functional connectivity, as well as the neural efficiency for experts' successful performance.

Altered Cortical Functional Networks in Patients With Schizophrenia and Bipolar Disorder: A Resting-State Electroencephalographic Study.

BACKGROUND: Pathologies of schizophrenia and bipolar disorder have been poorly understood. Brain network analysis could help understand brain mechanisms of schizophrenia and bipolar disorder. This study investigates the source-level brain cortical networks using resting-state electroencephalography (EEG) in patients with schizophrenia and bipolar disorder. METHODS: Resting-state EEG was measured in 38 patients with schizophrenia, 34 patients with bipolar disorder type I, and 30 healthy controls. Graph theory based source-level weighted functional networks were evaluated: strength, clustering coefficient (CC), path length (PL), and efficiency in six frequency bands. RESULTS: At the global level, patients with schizophrenia or bipolar disorder showed higher strength, CC, and efficiency, and lower PL in the theta band, compared to healthy controls. At the nodal level, patients with schizophrenia or bipolar disorder showed higher CCs, mostly in the frontal lobe for the theta band. Particularly, patients with schizophrenia showed higher nodal CCs in the left inferior frontal cortex and the left ascending ramus of the lateral sulcus compared to patients with bipolar disorder. In addition, the nodal-level theta band CC of the superior frontal gyrus and sulcus (cognition-related region) correlated with positive symptoms and social and occupational functioning scale (SOFAS) scores in the schizophrenia group, while that of the middle frontal gyrus (emotion-related region) correlated with SOFAS scores in the bipolar disorder group. CONCLUSIONS: Altered cortical networks were revealed and these alterations were significantly correlated with core pathological symptoms of schizophrenia and bipolar disorder. These source-level cortical network indices could be promising biomarkers to evaluate patients with schizophrenia and bipolar disorder.

Riemannian classifier enhances the accuracy of machine-learning-based diagnosis of PTSD using resting EEG.

Recently, objective and automated methods for the diagnosis of post-traumatic stress disorder (PTSD) have attracted increasing attention. However, previous studies on machine-learning-based diagnosis of PTSD with resting-state electroencephalogram (EEG) have reported poor accuracies of as low as 60%. Here, a Riemannian geometry-based classifier, the Fisher geodesic minimum distance to the mean (FgMDM), was employed for PTSD classification for the first time. Eyes-closed resting-state EEG data of 39 healthy individuals and 42 PTSD patients were used for the analysis. EEG source activities in 148 cortical regions were parcellated based on the Destrieux atlas, and their covariances were evaluated for each individual. Thirty epochs of preprocessed EEG were employed to calculate source activities. In addition, the FgMDM approach was applied to each EEG source covariance to construct the classifier. For a comparison, linear discriminant analysis (LDA), support vector machine (SVM), and random forest (RF) classifiers employing source band powers and network features as feature candidates were also tested. The FgMDM classifier showed an average classification accuracy of 75.240.80%. In contrast, the maximum accuracies of LDA, SVM, and RF classifiers were 66.54 ± 2.99%, 61.11 ± 2.98%, and 60.99 ± 2.19%, respectively. Our study demonstrated that the diagnostic accuracy of PTSD with resting-state EEG could be significantly improved by employing the FgMDM framework, which is a type of Riemannian geometry-based classifier.

Changes in Global and Nodal Networks in Patients With Unipolar Depression After 3-Week Repeated Transcranial Magnetic Stimulation Treatment.

Objectives: Repeated transcranial magnetic stimulation (rTMS) therapy has been applied in depressive disorders, but its neurobiological effect has not been well understood. Changes in cortical source network after treatment need to be confirmed. The present study investigated the effect of 3-week rTMS therapy on the symptom severity and cortical source network in patients with unipolar depression. Methods: Thirty-five patients with unipolar major depressive disorder participated in the study. High-frequency (10 Hz) rTMS was applied at the left dorsolateral prefrontal cortex during 3 weeks (five consecutive weekdays every week). Clinical symptoms were examined using the Hamilton Rating Scale for Depression and Anxiety. The resting state electroencephalography was recorded with 62 scalp channels before and after rTMS treatment. Results: Clinical symptoms significantly improved after rTMS treatment in both the active (p = 0.001) and sham groups (p = 0.002). However, an increased cortical source network in global and nodal levels was observed only in the active group after a 3-week treatment. Conclusions: The present study indicates that rTMS treatment leads to improved symptoms in patients with unipolar depression. Furthermore, treatment outcome of real effect was assured in changes of cortical source network.

Machine-learning-based classification between post-traumatic stress disorder and major depressive disorder using P300 features.

BACKGROUND: The development of optimal classification criteria for specific mental disorders which share similar symptoms is an important issue for precise diagnosis. We investigated whether P300 features in both sensor-level and source-level could be effectively used to classify post-traumatic stress disorder (PTSD) and major depressive disorder (MDD). METHOD: EEG signals were recorded from fifty-one PTSD patients, 67 MDD patients, and 39 healthy controls (HCs) while performing an auditory oddball task. Amplitude and latency of P300 were evaluated, and the current source analysis of P300 components was conducted using sLORETA. Finally, we classified two groups using machine-learning methods with both sensor- and source-level features. Moreover, we checked the comorbidity effects using the same approaches (PTSD-mono diagnosis (PTSDm, n = 28) and PTSD-comorbid diagnosis (PTSDc, n = 23)). RESULTS: PTSD showed significantly reduced P300 amplitudes and prolonged latency compared to HCs and MDD. Moreover, PTSD showed significantly reduced source activities, and the source activities were significantly correlated with symptoms of depression and anxiety. Also, the best classification accuracy at each pair was as follows: 80.00% (PTSD-HCs), 67.92% (MDD-HCs), 70.34% (PTSD-MDD), 82.09% (PTSDm-HCs), 71.58% (PTSDm-MDD), 82.56% (PTSDc-HCs), and 76.67% (PTSDc- MDD). CONCLUSION: Since abnormal P300 reflects pathophysiological characteristics of PTSD, PTSD patients were well-discriminated from MDD and HCs when using P300 features. Thus, altered P300 characteristics in both sensor- and source-level may be useful biomarkers to diagnosis PTSD.

Serum Levels of Tumor Necrosis Factor-α and Loudness Dependence of Auditory Evoked Potentials at Pretreatment and Posttreatment in Patients with Major Depressive Disorder.

BACKGROUND: Proinflammatory cytokines, such as tumor necrosis factor-alpha (TNF-α), are associated with the pathophysiology of major depressive disorder (MDD). Several studies have reported that increased TNF-α might be associated with tryptophan depletion, which eventually could result in MDD. However, other studies revealed that TNF-α increased serotonin firing in raphe. Therefore, whether TNF-α increases or decreases serotonin activity remains unclear. Here, we aimed to determine the relationship between serum TNF-α level and central serotonergic activity using the loudness dependence of auditory evoked potentials (LDAEP) and standardized low-resolution brain electromagnetic tomography (sLORETA), as well as to evaluate the effects of antidepressants on TNF-α levels. METHODS: LDAEP, serum TNF-α level, and depression severity were measured in 64 MDD outpatients pre and post 3 months of treatment. RESULTS: Pretreatment TNF-α levels were negatively correlated with the pretreatment N1 sLORETA-LDAEP, P2 sLORETA-LDAEP, and N1/P2 sLORETA-LDAEP (p < 0.05). In multiple regression analysis for N1/P2 sLORETA-LDAEP, lower N1/P2 sLORETA-LDAEP was significantly related to higher TNF-α (CE = -0.047, p = 0.017) when all subjects were dichotomized based on the median TNF-α level (7.16 pg/mL) into pretreatment low- and high-TNF-α groups. In addition, the pretreatment Beck Depression Inventory, P2 LDAEP, and N1/P2 sLORETA-LDAEP were greater in the high-TNF-α groups than in the low-TNF-α groups (p < 0.05). Moreover, the posttreatment TNF-α level was significantly decreased compared to the pretreatment TNF-α level (z = -2.581, p = 0.01). However, the posttreatment TNF-α levels were not associated with posttreatment LDAEP. CONCLUSIONS: Higher TNF-α level is associated with decreased LDAEP, which could indicate compensatory elevation of central serotonin activity in outpatients with MDD, although this effect disappeared and TNF-α level was reduced after three months of antidepressant treatment.