Review on EEG and ERP predictive biomarkers for major depressive disorder

One size does not fit all: notable individual variation in brain activity correlates of antidepressant treatment response

Introduction

To date, no robust electroencephalography (EEG) markers of antidepressant treatment response have been identified. Variable findings may arise from the use of group analyses, which neglect individual variation. Using a combination of group and single-participant analyses, we explored individual variability in EEG characteristics of treatment response.

Methods

Resting-state EEG data and Montgomery-Åsberg Depression Rating Scale (MADRS) symptom scores were collected from 43 patients with depression before, at 1 and 12 weeks of pharmacotherapy. Partial least squares (PLS) was used to: 1) identify group differences in EEG connectivity (weighted phase lag index) and complexity (multiscale entropy) between eventual medication responders and non-responders, and 2) determine whether group patterns could be identified in individual patients.

Results

Responders showed decreased alpha and increased beta connectivity, and early, widespread decreases in complexity over treatment. Non-responders showed an opposite connectivity pattern, and later, spatially confined decreases in complexity. Thus, as in previous studies, our group analyses identified significant differences between groups of patients with different treatment outcomes. These group-level EEG characteristics were only identified in ~40-60% of individual patients, as assessed quantitatively by correlating the spatiotemporal brain patterns between groups and individual results, and by independent raters through visualization.

Discussion

Our single-participant analyses suggest that substantial individual variation exists, and needs to be considered when investigating characteristics of antidepressant treatment response for potential clinical applicability.

Clinical trial registration

https://clinicaltrials.gov, identifier NCT00519428.

P300 parameters in major depressive disorder: A systematic review and meta-analysis

OBJECTIVES

Event-related potential measures have been extensively studied in mental disorders. Among them, P300 amplitude and latency reflect impaired cognitive abilities in major depressive disorder (MDD). The present systematic review and meta-analysis was conducted to investigate whether patients with MDD differ from healthy controls (HCs) with respect to P300 amplitude and latency.

METHODS

PubMed and Web of Science databases were searched from inception to 15 January 2023 for case-control studies comparing P300 amplitude and latency in patients with MDD and HCs. The primary outcome was the standard mean difference. A total of 13 articles on P300 amplitude and latency were included in the meta-analysis.

RESULTS

Random effect models indicated that MDD patients had decreased P300 amplitude, but similar latency compared to healthy controls. According to regression analysis, the effect size increased with the severity of depression and decreased with the proportion of women in the MDD samples. Funnel plot asymmetry was not significant for publication bias.

CONCLUSIONS

Decreased P300 amplitude may be a candidate diagnostic biomarker for MDD. However, prospective studies testing P300 amplitude as a monitoring biomarker for MDD are needed.

Electrophysiological biomarkers in dual pathology

INTRODUCTION

The co-occurrence of substance use disorder with at least one other mental disorder is called dual pathology, which in turn is characterised by heterogeneous symptoms that are difficult to diagnose and have a poor response to treatment. For this reason, the identification and validation of biomarkers is necessary. Within this group, possible electroencephalographic biomarkers have been reported to be useful in diagnosis, treatment and follow-up, both in neuropsychiatric conditions and in substance use disorders. This article aims to review the existing literature on electroencephalographic biomarkers in dual pathology.

METHODS

A narrative review of the literature. A bibliographic search was performed on the PubMed, Science Direct, OVID, BIREME and Scielo databases, with the keywords: electrophysiological biomarker and substance use disorder, electrophysiological biomarker and mental disorders, biomarker and dual pathology, biomarker and substance use disorder, electroencephalography, and substance use disorder or comorbid mental disorder.

RESULTS

Given the greater amount of literature found in relation to electroencephalography as a biomarker of mental illness and substance use disorders, and the few articles found on dual pathology, the evidence is organised as a biomarker in psychiatry for the diagnosis and prediction of risk and as a biomarker for dual pathology.

CONCLUSIONS

Although the evidence is not conclusive, it suggests the existence of a subset of sites and mechanisms where the effects of psychoactive substances and the neurobiology of some mental disorders could overlap or interact.

Neural energy computations based on Hodgkin-Huxley models bridge abnormal neuronal activities and energy consumption patterns of major depressive disorder

Limited by the current experimental techniques and neurodynamical models, the dysregulation mechanisms of decision-making related neural circuits in major depressive disorder (MDD) are still not clear. In this paper, we proposed a neural coding methodology using energy to further investigate it, which has been proven to strongly complement the neurodynamical methodology. We augmented the previous neural energy calculation method, and applied it to our VTA-NAc-mPFC neurodynamical H-H models. We particularly focused on the peak power and energy consumption of abnormal ion channel (ionic) currents under different concentrations of dopamine input, and investigated the abnormal energy consumption patterns for the MDD group. The results revealed that the energy consumption of medium spiny neurons (MSNs) in the NAc region were lower in the MDD group than that of the normal control group despite having the same firing frequencies, peak action potentials, and average membrane potentials in both groups. Dopamine concentration was also positively correlated with the energy consumption of the pyramidal neurons, but the patterns of different interneuron types were distinct. Additionally, the ratio of mPFC's energy consumption to total energy consumption of the whole network in MDD group was lower than that in normal control group, revealing that the mPFC region in MDD group encoded less neural information, which matched the energy consumption patterns of BOLD-fMRI results. It was also in line with the behavioral characteristics that MDD patients demonstrated in the form of reward insensitivity during decision-making tasks. In conclusion, the model in this paper was the first neural network energy computational model for MDD, which showed success in explaining its dynamical mechanisms with an energy consumption perspective. To build on this, we demonstrated that energy consumption levels can be used as a potential indicator for MDD, which also showed a promising pipeline to use an energy methodology for studying other neuropsychiatric disorders.

Electrophysiological phenotypes of suicidality predict prolonged response to oral ketamine treatment

Oral ketamine has shown to be a rapid-acting antidepressant and a potential treatment option for suicidality, however, repeated doses are often required. Objective markers of prolonged treatment response are needed to help individuals and clinicians make informed treatment decisions. This secondary analysis sought to identify objective electrophysiological predictors of both prolonged response and dose sensitivity to low-dose oral ketamine in people with chronic suicidality. Individuals with a Beck Scale for Suicide Ideation total score (BSS) ≥ 6 (N = 29) completed a six-week ketamine treatment, pre-treatment electroencephalography and follow-up assessment of suicidality (four weeks from the final ketamine dose). Prolonged response was observed in 52% of participants (follow-up BSS reduced by 50% or ≤6); nearly half were prolonged non-responders. There was decisive evidence for a predictive Bayesian linear regression model with follow-up BSS score as the response variable and pre-treatment auditory evoked power bands as predictors (theta, alpha and beta frequencies, BF₁₀ = 17,948, R² = 0.70). A Bayesian one-way ANOVA indicated strong evidence for a model of positive association between auditory evoked power and ketamine dose sensitivity (theta-alpha BF₊₀ = 108, effect size δ = 1.3, 95% CI 0.5-2.1; high-beta BF₊₀ = 7.4, δ = 0.8, 95% CI 0.1-1.6). Given auditory evoked power may index serotonin neurotransmission, these results suggest that a prolonged response to ketamine may, in part, be mediated by pre-treatment serotonergic functioning. In addition, the observed beta power differences may arise from GABAergic functioning. These suicidality phenotypes, identifiable by pre-treatment electrophysiology, may aid diagnosis, treatment selection and prediction of prolonged treatment outcome.

EEG frequency band analysis in chronic neuropathic pain: A linear and nonlinear approach to classify pain severity

BACKGROUND AND OBJECTIVE

Chronic neuropathic pain (NP) is a chronic pain condition that severely impacts a patient's life. Pain management has proved to be inefficient due to a lack of a simple clinical tool that may identify and monitor NP. A low-cost, noninvasive tool that provides relevant information on NP is the electroencephalogram (EEG). However, the commonly used linear EEG features have proved to be limited in characterizing NP pathophysiology. This study sought to determine whether nonlinear EEG features such as approximate entropy (ApEn) would better differentiate pain severity than absolute band power.

METHODS

A non-parametric statistical approach based on the Brief Pain Inventory (BPI), along with linear and nonlinear EEG features, is proposed in this study. For this purpose, thirty-six chronic NP patients were recruited, and 22 channels were registered. Additionally, a control database of 13 participants with no NP was used as a reference, where 19 channels were registered. For both groups, EEG was recorded for 10 min in a resting state: 5 min with eyes open (EO) and 5 min with eyes closed (EC). Absolute band power and ApEn EEG features in the five clinical frequency bands (delta, theta, alpha, beta, and gamma) were estimated for all channels in both groups. As a result, 220-dimensional and 190-dimensional feature vectors were obtained for experimental and control classes respectively. For the experimental class, NP patients were grouped according to their BPI evaluation in three groups: low, moderate, and high pain. Finally, feature vectors were compared between groups using Kruskal Wallis and post-hoc Dunn's tests.

RESULTS

ApEn revealed significant statistical difference (p <=0.0001) in most frequency bands and conditions among the groups. In contrast, power had less significant differences between groups, particularly with EO. Furthermore, NP groups were notably clustered using only ApEn in theta, alpha, and beta bands.

CONCLUSIONS

The results indicate that ApEn effectively characterizes the different severities of chronic NP rather than the commonly used linear features. ApEn and other nonlinear techniques (e.g., spectral entropy, Shannon entropy) might be a more suitable methodology to monitor chronic NP experience.

Electrophysiological biomarkers and age characterize phenotypic heterogeneity among individuals with major depressive disorder

Introduction: Despite the high need for effective treatments for major depressive disorder (MDD), the development of novel medicines is hampered by clinical, genetic and biological heterogeneity, unclear links between symptoms and neural dysfunction, and tenuous biomarkers for clinical trial contexts of use. Methods: In this study, we examined the International Study to Predict Optimized Treatment in Depression (iSPOT-D) clinical trial database for new relationships between auditory event-related potential (ERP) responses, demographic features, and clinical symptoms and behavior, to inform strategies for biomarker-driven patient stratification that could be used to optimize future clinical trial design and drug development strategy in MDD. Results: We replicate findings from previous analyses of the classic auditory oddball task in the iSPOT-D sample showing smaller than typical N1 and P300 response amplitudes and longer P300 latencies for target and standard stimuli in patients with MDD, suggesting altered bottom-up sensory and top-down attentional processes. We further demonstrate that age is an important contributor to clinical group differences, affecting both topographic distribution of the clinically informative ERP responses and the types of the stimuli sensitive to group differences. In addition, the observed brain-behavior associations indicate that levels of anxiety and stress are major contributing factors to atypical sensory and attentional processing among patients with MDD, particularly in the older subgroups. Discussion: Our novel findings support the possibility of accelerated cognitive aging in patients with MDD and identify the frontal P300 latency as an additional candidate biomarker of MDD. These results from a large, well-phenotyped sample support the view that heterogeneity of the clinical population with MDD can be systematically characterized based on age and neural biomarkers of sensory and attentional processing, informing patient stratification strategies in the design of clinical trials.

Detecting depression of Chinese microblog users via text analysis: Combining Linguistic Inquiry Word Count (LIWC) with culture and suicide related lexicons

INTRODUCTION

In recent years, research has used psycholinguistic features in public discourse, networking behaviors on social media and profile information to train models for depression detection. However, the most widely adopted approach for the extraction of psycholinguistic features is to use the Linguistic Inquiry Word Count (LIWC) dictionary and various affective lexicons. Other features related to cultural factors and suicide risk have not been explored. Moreover, the use of social networking behavioral features and profile features would limit the generalizability of the model. Therefore, our study aimed at building a prediction model of depression for text-only social media data through a wider range of possible linguistic features related to depression, and illuminate the relationship between linguistic expression and depression.

METHODS

We collected 789 users' depression scores as well as their past posts on Weibo, and extracted a total of 117 lexical features via Simplified Chinese Linguistic Inquiry Word Count, Chinese Suicide Dictionary, Chinese Version of Moral Foundations Dictionary, Chinese Version of Moral Motivation Dictionary, and Chinese Individualism/Collectivism Dictionary.

RESULTS

Results showed that all the dictionaries contributed to the prediction. The best performing model occurred with linear regression, with the Pearson correlation coefficient between predicted values and self-reported values was 0.33, the R-squared was 0.10, and the split-half reliability was 0.75.

DISCUSSION

This study did not only develop a predictive model applicable to text-only social media data, but also demonstrated the importance taking cultural psychological factors and suicide related expressions into consideration in the calculation of word frequency. Our research provided a more comprehensive understanding of how lexicons related to cultural psychology and suicide risk were associated with depression, and could contribute to the recognition of depression.

Decoding of Processing Preferences from Language Paradigms by Means of EEG-ERP Methodology: Risk Markers of Cognitive Vulnerability for Depression and Protective Indicators of Well-Being? Cerebral Correlates and Mechanisms

Depression is a frequent mental affective disorder. Cognitive vulnerability models propose two major cognitive risk factors that favor the onset and severity of depressive symptoms. These include a pronounced self-focus, as well as a negative emotional processing bias. According to two-process models of cognitive vulnerability, these two risk factors are not independent from each other, but affect information processing already at an early perceptual processing level. Simultaneously, a processing advantage for self-related positive information including better memory for positive than negative information has been associated with mental health and well-being. This perspective paper introduces a research framework that discusses how EEG-ERP methodology can serve as a standardized tool for the decoding of negative and positive processing biases and their potential use as risk markers of cognitive vulnerability for depression, on the one hand, and as protective indicators of well-being, on the other hand. Previous results from EEG-ERP studies investigating the time-course of self-referential emotional processing are introduced, summarized, and discussed with respect to the specificity of depression-related processing and the importance of EEG-ERP-based experimental testing for well-being and the prevention and treatment of depressive disorders.

Auditory event-related potentials in separating patients with depressive disorders and non-depressed controls: A narrative review

This narrative review brings together the findings regarding the differences in the auditory event-related potentials (ERPs) between patients with depressive disorder and non-depressed control subjects. These studies' results can inform us of the possible alterations in sensory-cognitive processing in depressive disorders and the potential of using these ERPs in clinical applications. Auditory P3, mismatch negativity (MMN) and loudness dependence of auditory evoked potentials (LDAEP) were the subjects of the investigation. A search in PubMed yielded 84 studies. The findings of the reviewed studies were not highly consistent, but some patterns could be identified. For auditory P3b, the common findings were attenuated amplitude and prolonged latency among depressed patients. Regarding auditory MMN, especially the amplitude of duration deviance MMN was commonly attenuated, and the amplitude of frequency deviance MMN was increased in depressed patients. In LDAEP studies, generally, no differences between depressed patients and non-depressed controls were reported, although some group differences concerning specific depression subtypes were found. This review posits that future research should investigate whether certain stimulus conditions are particularly efficient at separating depressed and non-depressed participant groups. Future studies should contrast responses in different subpopulations of depressed patients, as well as different clinical groups (e.g., depressive disorder and anxiety disorder patients), to investigate the specificity of the auditory ERP alterations for depressive disorders.

Detecting and Measuring Depression on Social Media Using a Machine Learning Approach: Systematic Review

BACKGROUND

Detection of depression gained prominence soon after this troublesome disease emerged as a serious public health concern worldwide.

OBJECTIVE

This systematic review aims to summarize the findings of previous studies concerning applying machine learning (ML) methods to text data from social media to detect depressive symptoms and to suggest directions for future research in this area.

METHODS

A bibliographic search was conducted for the period of January 1990 to December 2020 in Google Scholar, PubMed, Medline, ERIC, PsycINFO, and BioMed. Two reviewers retrieved and independently assessed the 418 studies consisting of 322 articles identified through database searching and 96 articles identified through other sources; 17 of the studies met the criteria for inclusion.

RESULTS

Of the 17 studies, 10 had identified depression based on researcher-inferred mental status, 5 had identified it based on users' own descriptions of their mental status, and 2 were identified based on community membership. The ML approaches of 13 of the 17 studies were supervised learning approaches, while 3 used unsupervised learning approaches; the remaining 1 study did not describe its ML approach. Challenges in areas such as sampling, optimization of approaches to prediction and their features, generalizability, privacy, and other ethical issues call for further research.

CONCLUSIONS

ML approaches applied to text data from users on social media can work effectively in depression detection and could serve as complementary tools in public mental health practice.

Cognitive assessment using ERP in child and adolescent psychiatry: Difficulties and opportunities

Event related potentials (ERPs) represent powerful tools to investigate cognitive functioning in child and adolescent psychiatry. So far, the available body of research has largely focused on advancements in analysis methods, with little attention given to the perspective of assessment. The aim of this brief report is to provide recommendations for cognitive ERPs assessment that can be applied across diagnostic categories in child and adolescent psychiatry research. First, we discuss major issues for ERPs testing using examples from common psychiatric disorders. We conclude by summing up our recommendations for methodological standards and highlighting the potential role of ERPs in the field.

An Eye Tracking and Event-Related Potentials Study With Visual Stimuli for Adolescents Emotional Issues

BACKGROUND

Psychological issues are common among adolescents, which have a significant impact on their growth and development. However, the underlying neural mechanisms of viewing visual stimuli in adolescents are poorly understood.

MATERIALS AND METHODS

This study applied the Chinese version of the DSM-V self-assessment scales to evaluate 73 adolescents' psychological characteristics for depressive and manic emotional issues. Combined with eye-tracking and event-related potential (ERP), we explored the characteristics of their visual attention and neural processing mechanisms while freely viewing positive, dysphoric, threatening and neutral visual stimuli.

RESULTS

Compared to controls, adolescents with depressive emotional tendencies showed more concentrated looking behavior with fixation distribution index than the controls, while adolescents with manic emotional tendencies showed no such trait. ERP data revealed individuals with depressive tendencies showed lower arousal levels toward emotional stimuli in the early stage of cognitive processing (N1 amplitude decreased) and with prolonged reaction time (N1 latency increased) than the control group. We found no significant difference between the manic group and the control group. Furthermore, the depression severity scores of the individuals with depressive tendencies were negatively correlated with the total fixation time toward positive stimuli, were negatively correlated with the fixation distribution index toward threatening stimuli, and were positively correlated with the mean N1 amplitudes while viewing dysphoric stimuli. Also, for the individuals with depressive tendencies, there was a positive correlation between the mean N1 amplitudes and the fixation time on the area of interest (AOI) while viewing dysphoric stimuli. For the individuals with manic tendencies, the manic severity scores of the individuals with manic tendencies were positively correlated with the total fixation time toward the positive stimuli. However, no significant correlations were found between the manic severity scores and N1 amplitudes, and between N1 amplitudes and eye-tracking output variables.

CONCLUSION

This study proposes the application of eye-tracking and ERP to provide better biological evidence to alter the neural processing of emotional stimuli for adolescents with emotional issues.

Feature Extraction to Identify Depression and Anxiety Based on EEG

Biomarkers in neurophysiological signals can be analyzed to determine indicators of mood disorders for diagnosis. In this paper, EEG signals were analyzed from a public database of 119 subjects ages 18 to 24 performing a cognitive task. 45 subjects had moderate to severe anxiety and/or depression and the remaining 74 subjects had minimal or none. A subject's level of depression and/or anxiety was classified by standard psychological tests. EEG signals were preprocessed and separated into frequency bands: beta (12-30 Hz), alpha (8-12 Hz), theta (4-8 Hz) and delta (0.5-4 Hz). Features were extracted including Higuchi Fractal Dimension, correlation dimension, approximate entropy, Lyapunov exponent and detrended fluctuation analysis. Similarities, and asymmetry can be examined between the left and right brain hemispheres as well as the prefrontal cortex channels. ANOVA II analysis showed a significant difference (p<0.05) for topographical region comparisons of several features between the affected and unaffected subjects for specific features. The results demonstrate physiological asymmetry between high scoring subjects indicating a mood disorder, with low scoring, to be used as an indicator of illness. Understanding the complexities of how depression and anxiety are manifested physiologically including its comorbidities, is critical for accurate and objective diagnosis of mood and anxiety order disorders.

Statistical Analysis of Graph-Theoretic Indices to Study EEG-TMS Connectivity in Patients With Depression

Aim

The objective of this work was to demonstrate the usefulness of a novel statistical method to study the impact of transcranial magnetic stimulation (TMS) on brain connectivity in patients with depression using different stimulation protocols, i.e., 1 Hz repetitive TMS over the right dorsolateral prefrontal cortex (DLPFC) (protocol G1), 10 Hz repetitive TMS over the left DLPFC (G2), and intermittent theta burst stimulation (iTBS) consisting of three 50 Hz burst bundle repeated at 5 Hz frequency (G3).

Methods

Electroencephalography (EEG) connectivity analysis was performed using Directed Transfer Function (DTF) and a set of 21 indices based on graph theory. The statistical analysis of graph-theoretic indices consisted of a combination of the k-NN rule, the leave-one-out method, and a statistical test using a 2 × 2 contingency table.

Results

Our new statistical approach allowed for selection of the best set of graph-based indices derived from DTF, and for differentiation between conditions (i.e., before and after TMS) and between TMS protocols. The effects of TMS was found to differ based on frequency band.

Conclusion

A set of four brain asymmetry measures were particularly useful to study protocol- and frequency-dependent effects of TMS on brain connectivity.

Significance

The new approach would allow for better evaluation of the therapeutic effects of TMS and choice of the most appropriate stimulation protocol.

Does mismatch negativity have utility for NMDA receptor drug development in depression?

CLINICAL TRIAL REGISTRATION

Rapid antidepressant effects associated with ketamine have shifted the landscape for the development of therapeutics to treat major depressive disorder (MDD) from a monoaminergic to glutamatergic model. Treatment with ketamine, an N-methyl-D-aspartate (NMDA) receptor antagonist, may be effective, but has many non-glutamatergic targets, and clinical and logistical problems are potential challenges. These factors underscore the importance of manipulations of binding mechanics to produce antidepressant effects without concomitant clinical side effects. This will require identification of efficient biomarkers to monitor target engagement. The mismatch negativity (MMN) is a widely used electrophysiological signature linked to the activity of NMDA receptors (NMDAR) in humans and animals and validated in pre-clinical and clinical studies of ketamine. In this review, we explore the flexibility of the MMN and its capabilities for reliable use in drug development for NMDAR antagonists in MDD. We supplement this with findings from our own research with three distinct NMDAR antagonists. The research described illustrates that there are important distinctions between the mechanisms of NMDAR antagonism, which are further crystallized when considering the paradigm used to study the MMN. We conclude that the lack of standardized methodology currently prevents MMN from being ready for common use in drug discovery. This manuscript describes data collected from the following National Institutes of Health (NIH) and Veterans Affairs (VA) studies: AV-101, NCT03583554; lanicemine, NCT03166501; ketamine, NCT02556606.

EEG based Major Depressive disorder and Bipolar disorder detection using Neural Networks:A review

Mental disorders represent critical public health challenges as they are leading contributors to the global burden of disease and intensely influence social and financial welfare of individuals. The present comprehensive review concentrate on the two mental disorders: Major depressive Disorder (MDD) and Bipolar Disorder (BD) with noteworthy publications during the last ten years. There is a big need nowadays for phenotypic characterization of psychiatric disorders with biomarkers. Electroencephalography (EEG) signals could offer a rich signature for MDD and BD and then they could improve understanding of pathophysiological mechanisms underling these mental disorders. In this review, we focus on the literature works adopting neural networks fed by EEG signals. Among those studies using EEG and neural networks, we have discussed a variety of EEG based protocols, biomarkers and public datasets for depression and bipolar disorder detection. We conclude with a discussion and valuable recommendations that will help to improve the reliability of developed models and for more accurate and more deterministic computational intelligence based systems in psychiatry. This review will prove to be a structured and valuable initial point for the researchers working on depression and bipolar disorders recognition by using EEG signals.

A reduced P300 prospectively predicts increased depressive severity in adults with clinical depression

Neurocognitive impairments commonly observed in depressive disorders are thought to be reflected in reduced P300 amplitudes. To date, depression-related P300 amplitude reduction has mostly been demonstrated cross-sectionally, while its clinical implication for the course of depression remains largely unclear. Moreover, the relationship between P300 and specific clinical characteristics of depression is uncertain. To shed light on the functional significance of the P300 in depression, we examined whether initial P300 amplitude prospectively predicted changes in depressive symptoms among a community sample of 58 adults (mean age = 38.86 years old, 81% female) with a current depressive disorder. This sample was assessed at two-time points, separated by approximately nine months (range = 6.6-15.9). At the initial visit, participants completed clinical interviews, self-report measures, and a flanker task, while EEG was recorded to derive P300 amplitude. At the follow-up visit, participants again completed the same clinical interviews and self-report measures. Results indicated that a reduced P300 amplitude at the initial visit was associated with higher total depressive symptoms at follow-up, even after controlling for initial depressive symptoms. These data indicate the potential clinical utility for the P300 as a neural marker of disease course among adults with a current depressive disorder. Future research may target P300 in interventions to determine whether depression-related outcomes can be improved.

Using EEG to Predict Clinical Response to Electroconvulsive Therapy in Patients With Major Depression: A Comprehensive Review

Introduction: An important approach to improve the therapeutic effect of electroconvulsive therapy (ECT) may be to early characterize patients who are more likely to respond. Our objective was to explore whether baseline electroencephalography (EEG) settings before the beginning of ECT treatment can predict future clinical response to ECT in patients with depressive disorder. Methods: We conducted a systematic search in the MEDLINE, EMBASE, PsycINFO, Web of Science, and Cochrane Central Register of Controlled Trials (CENTRAL) databases to identify studies using EEG in adults with depressive disorder treated by ECT. To investigate the predictive value of baseline EEG on clinical outcomes of ECT, we extracted from the retrieved studies and qualitatively described the association between the baseline EEG markers characteristics and the rates of future responders and/or remitters to ECT. Results: The primary search yielded 2,531 potentially relevant citations, and 12 articles were selected according to inclusion criteria. Most of the studies were prospective studies with small sample size. Sociodemographic and clinical characteristics of patients, ECT settings, EEG settings, and outcomes were heterogeneous. Event-related potential (ERP) paradigms were used in three studies, polysomnography was used in three studies, and the six other studies used EEG to measure cerebral connectivity and activity. Conclusions: P300 amplitude, coherence, and connectivity measures were correlated with remission in patients with depression treated by ECT. Sleep EEG recordings seemed not to be correlated with remission after ECT. Further prospective studies with large sample size are needed to determine optimal EEG parameters associated with clinical response to ECT in depressive disorder. Systematic Review Registration: PROSPERO CRD42020181978.

Major Depression and Brain Asymmetry in a Decision-Making Task with Negative and Positive Feedback

Depressed patients are characterized by hypoactivity of the left and hyperactivity of the right frontal areas during the resting state. Depression is also associated with impaired decision-making, which reflects multiple cognitive, affective, and attentional processes, some of which may be lateralized. The aim of this study was to investigate brain asymmetry during a decision-making task performed in negative and positive feedback conditions in patients with Major Depressive Disorder (MDD) in comparison to healthy control participants. The electroencephalogram (EEG) was recorded from 60 MDD patients and 60 healthy participants while performing a multi-stage decision-making task. Frontal, central, and parietal alpha asymmetry were analyzed with EEGlab/ERPlab software. Evoked potential responses (ERPs) showed general lateralization suggestive of an initial right dominance developing into a more complex pattern of asymmetry across different scalp areas as information was processed. The MDD group showed impaired mood prior to performance, and decreased confidence during performance in comparison to the control group. The resting state frontal alpha asymmetry showed lateralization in the healthy group only. Task-induced alpha power and ERP P100 and P300 amplitudes were more informative biomarkers of depression during decision making. Asymmetry coefficients based on task alpha power and ERP amplitudes showed consistency in the dynamical changes during the decision-making stages. Depression was characterized by a lack of left dominance during the resting state and left hypoactivity during the task baseline and subsequent decision-making process. Findings add to understanding of the functional significance of lateralized brain processes in depression.

Functional connectivity of major depression disorder using ongoing EEG during music perception

OBJECTIVE

The functional connectivity (FC) of major depression disorder (MDD) has not been well studied under naturalistic and continuous stimuli conditions. In this study, we investigated the frequency-specific FC of MDD patients exposed to conditions of music perception using ongoing electroencephalogram (EEG).

METHODS

First, we applied the phase lag index (PLI) method to calculate the connectivity matrices and graph theory-based methods to measure the topology of brain networks across different frequency bands. Then, classification methods were adopted to identify the most discriminate frequency band for the diagnosis of MDD.

RESULTS

During music perception, MDD patients exhibited a decreased connectivity pattern in the delta band but an increased connectivity pattern in the beta band. Healthy people showed a left hemisphere-dominant phenomenon, but MDD patients did not show such a lateralized effect. Support vector machine (SVM) achieved the best classification performance in the beta frequency band with an accuracy of 89.7%, sensitivity of 89.4% and specificity of 89.9%.

CONCLUSIONS

MDD patients exhibited an altered FC in delta and beta bands, and the beta band showed a superiority in the diagnosis of MDD.

SIGNIFICANCE

Our study provided a promising reference for the diagnosis of MDD, and revealed a new perspective for understanding the topology of MDD brain networks during music perception.

Prediction of Antidepressant Treatment Outcome Using Event-Related Potential in Patients with Major Depressive Disorder

(1) Background: Prediction of treatment outcome has been one of the core objectives in clinical research of patients with major depressive disorder (MDD). This study explored the possibility of event-related potential (ERP) markers to predict antidepressant treatment outcomes among MDD patients; (2) Methods: Fifty-two patients with MDD were recruited and evaluated through Hamilton depression (HAM-D), Hamilton anxiety rating scale (HAM-A), and CORE. Patients underwent a battery of ERP measures including frontal alpha symmetry (FAA) in the low alpha band (8–10 Hz), mismatch negativity (MMN), and loudness-dependent auditory evoked potentials (LDAEP); (3) Results: During the eight weeks of study, 61% of patients achieved remission, and 77% showed successful treatment responsiveness. Patients with low FAA in F5/F6 demonstrated a significantly higher remission/response ratio and better treatment responsiveness (F (2.560, 117.755) = 3.84, p = 0.016) compared to patients with high FAA. In addition, greater FAA in F7/F8 EEG channels was significantly associated with greater melancholia scores (r = 0.34, p = 0.018). Other ERP markers lacked any significant effect; (4) Conclusions: Our results suggested low FAA (i.e., greater left frontal activity) could reflect a good treatment response in MDD patients. These findings support that FAA could be a promising index in understanding both MDD and melancholic subtype.

Cortical Connectivity Moderators of Antidepressant vs Placebo Treatment Response in Major Depressive Disorder: Secondary Analysis of a Randomized Clinical Trial

IMPORTANCE

Despite the widespread awareness of functional magnetic resonance imaging findings suggesting a role for cortical connectivity networks in treatment selection for major depressive disorder, its clinical utility remains limited. Recent methodological advances have revealed functional magnetic resonance imaging-like connectivity networks using electroencephalography (EEG), a tool more easily implemented in clinical practice.

OBJECTIVE

To determine whether EEG connectivity could reveal neural moderators of antidepressant treatment.

DESIGN, SETTING, AND PARTICIPANTS

In this nonprespecified secondary analysis, data were analyzed from the Establishing Moderators and Biosignatures of Antidepressant Response in Clinic Care study, a placebo-controlled, double-blinded randomized clinical trial. Recruitment began July 29, 2011, and was completed December 15, 2015. A random sample of 221 outpatients with depression aged 18 to 65 years who were not taking medication for depression was recruited and assessed at 4 clinical sites. Analysis was performed on an intent-to-treat basis. Statistical analysis was performed from November 16, 2018, to May 23, 2019.

INTERVENTIONS

Patients received either the selective serotonin reuptake inhibitor sertraline hydrochloride or placebo for 8 weeks.

MAIN OUTCOMES AND MEASURES

Electroencephalographic orthogonalized power envelope connectivity analyses were applied to resting-state EEG data. Intent-to-treat prediction linear mixed models were used to determine which pretreatment connectivity patterns were associated with response to sertraline vs placebo. The primary clinical outcome was the total score on the 17-item Hamilton Rating Scale for Depression, administered at each study visit.

RESULTS

Of the participants recruited, 9 withdrew after first dose owing to reported adverse effects, and 221 participants (150 women; mean [SD] age, 37.8 [12.7] years) underwent EEG recordings and had high-quality pretreatment EEG data. After correction for multiple comparisons, connectome-wide analyses revealed moderation by connections within and between widespread cortical regions-most prominently parietal-for both the antidepressant and placebo groups. Greater alpha-band and lower gamma-band connectivity predicted better placebo outcomes and worse antidepressant outcomes. Lower connectivity levels in these moderating connections were associated with higher levels of anhedonia. Connectivity features that moderate treatment response differentially by treatment group were distinct from connectivity features that change from baseline to 1 week into treatment. The group mean (SD) score on the 17-item Hamilton Rating Scale for Depression was 18.35 (4.58) at baseline and 26.14 (30.37) across all time points.

CONCLUSIONS AND RELEVANCE

These findings establish the utility of EEG-based network functional connectivity analyses for differentiating between responses to an antidepressant vs placebo. A role emerged for parietal cortical regions in predicting placebo outcome. From a treatment perspective, capitalizing on the therapeutic components leading to placebo response differentially from antidepressant response should provide an alternative direction toward establishing a placebo signature in clinical trials, thereby enhancing the signal detection in randomized clinical trials.

TRIAL REGISTRATION

ClinicalTrials.gov identifier: NCT01407094.

Differential effects of traxoprodil and S-ketamine on quantitative EEG and auditory event-related potentials as translational biomarkers in preclinical trials in rats and mice

Quantitative Electroencephalography (qEEG) and event-related potential (ERP) assessment have emerged as powerful tools to unravel translational biomarkers in preclinical and clinical psychiatric drug discovery trials. The aim of the present study was to compare the GluN2B negative allosteric modulator (NAM) traxoprodil (CP-101,606) with the unselective NMDA receptor channel blocker S-ketamine to give insight into central target engagement and differentiation on multiple EEG readouts. For qEEG recordings telemetric transmitters were implanted in male Wistar rats. Recorded EEG data were analyzed using fast Fourier transformation to determine power spectra and vigilance states. Additionally, body temperature and locomotor activity were assessed via telemetry. For recordings of auditory event-related potentials (AERP) male C57Bl/6J mice were chronically implanted with deep electrodes using a tethered system. Power spectral analysis revealed a significant increase in gamma power following ketamine treatment, whereas traxoprodil (6&18 mg/kg) induced an overall decrease primarily within alpha and beta bands. Additionally, ketamine disrupted sleep and enhanced time spent in wake vigilance states, whereas traxoprodil did not alter sleep-wake architecture. AERP and mismatch negativity (MMN) revealed that ketamine (10 mg/kg) selectively disrupts auditory deviance detection, whereas traxoprodil (6 mg/kg) did not alter MMN at clinically relevant doses. In contrast to ketamine treatment, traxoprodil did not produce hyperactivity and hypothermia. In conclusion, ketamine and traxoprodil showed very different effects on diverse EEG readouts differentiating selective GluN2B antagonism from non-selective pan-NMDA-R antagonists like ketamine. These readouts are thus perfectly suited to support drug discovery efforts on NMDA-R and understanding the different functions of NMDA-R subtypes.

Effects of working memory training on EEG, cognitive performance, and self-report indices potentially relevant for social anxiety

Social anxiety (SA) is quite common and associated with multiple comorbidities. Here, we examined the effects of working memory (WM) training on various indices potentially related to SA. Pre-selected university students with elevated self-reported SA symptoms were assigned to a WM training (n = 21) or an active control treatment condition (n = 21). Pre- and post-treatment assessments were made using questionnaires related to (social) anxiety and depression, and tasks measuring WM, interference control, and attentional biases towards, and event-related potentials (ERPs) elicited by, angry faces. The training enhanced WM transfer task performance, reduced SA symptoms, and changed the amplitude of the P1, N170, P2, and N2 ERP components. However, the latter changes did not mediate the effect of WM training on SA symptoms. These data provide preliminary evidence of the usefulness of WM trainings to reduce potential indices of SA, but further research is necessary to unravel the causal relation among these indices.

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.

Differentiating responders and non-responders to rTMS treatment for depression after one week using resting EEG connectivity measures

BACKGROUND

Non-response to repetitive transcranial magnetic stimulation (rTMS) treatment for depression is costly for both patients and clinics. Simple and cheap methods to predict response would reduce this burden. Resting EEG measures differentiate responders from non-responders, so may have utility for response prediction.

METHODS

Fifty patients with treatment resistant depression and 21 controls had resting electroencephalography (EEG) recorded at baseline (BL). Patients underwent 5-8 weeks of rTMS treatment, with EEG recordings repeated at week 1 (W1). Forty-two participants had valid BL and W1 EEG data, and 12 were responders. Responders and non-responders were compared at BL and W1 in measures of theta (4-8 Hz) and alpha (8-13 Hz) power and connectivity, frontal theta cordance and alpha peak frequency. Control group comparisons were made for measures that differed between responders and non-responders. A machine learning algorithm assessed the potential to differentiate responders from non-responders using EEG measures in combination with change in depression scores from BL to W1.

RESULTS

Responders showed elevated theta connectivity across BL and W1. No other EEG measures differed between groups. Responders could be distinguished from non-responders with a mean sensitivity of 0.84 (p = 0.001) and specificity of 0.89 (p = 0.002) using cross-validated machine learning classification on the combination of all EEG and mood measures.

LIMITATIONS

The low response rate limited our sample size to only 12 responders.

CONCLUSION

Resting theta connectivity at BL and W1 differ between responders and non-responders, and show potential for predicting response to rTMS treatment for depression.

Depression Detection Using Relative EEG Power Induced by Emotionally Positive Images and a Conformal Kernel Support Vector Machine

Electroencephalography (EEG) can assist with the detection of major depressive disorder (MDD). However, the ability to distinguish adults with MDD from healthy individuals using resting-state EEG features has reached a bottleneck. To address this limitation, we collected EEG data as participants engaged with positive pictures from the International Affective Picture System. Because MDD is associated with blunted positive emotions, we reasoned that this approach would yield highly dissimilar EEG features in healthy versus depressed adults. We extracted three types of relative EEG power features from different frequency bands (delta, theta, alpha, beta, and gamma) during the emotion task and resting state. We also applied a novel classifier, called a conformal kernel support vector machine (CK-SVM), to try to improve the generalization performance of conventional SVMs. We then compared CK-SVM performance with three machine learning classifiers: linear discriminant analysis (LDA), conventional SVM, and quadratic discriminant analysis. The results from the initial analyses using the LDA classifier on 55 participants (24 MDD, 31 healthy controls) showed that the participant-independent classification accuracy obtained by leave-one-participant-out cross-validation (LOPO-CV) was higher for the EEG recorded during the positive emotion induction versus the resting state for all types of relative EEG power. Furthermore, the CK-SVM classifier achieved higher LOPO-CV accuracy than the other classifiers. The best accuracy (83.64%; sensitivity = 87.50%, specificity = 80.65%) was achieved by the CK-SVM, using seven relative power features extracted from seven electrodes. Overall, combining positive emotion induction with the CK-SVM classifier proved useful for detecting MDD on the basis of EEG signals. In the future, this approach might be used to develop a brain–computer interface system to assist with the detection of MDD in the clinic. Importantly, such a system could be implemented with a low-density electrode montage (seven electrodes), highlighting its practical utility.

Event related potential analysis techniques for autism spectrum disorders: A review

Autism Spectrum Disorders (ASD) comprise all pervasive neurodevelopmental diseases marked by deficits in social and communication skills, delayed cognitive development, restricted and repetitive behaviors. The core symptoms begin in early childhood, may continue life-long resulting in poor performance in adult stage. Event-related potential (ERP) is basically a time-locked electroencephalogram signal elicited by various stimuli, related to sensory and cognitive processes. The various ERP based techniques used for the study of ASD are considered in this review. ERP based study offers the advantage of being a non-invasive technique to measure the brain activity precisely. The techniques are categorized into three based on the processing domain: time, frequency and time-frequency. Power spectral density, coherence, phase synchrony, multiscale entropy, modified multiscale entropy, sum of signed differences, synchrostates and variance are some of the measures that have been widely used to study the abnormalities in frequency bands and brain connectivity. Various signal processing techniques such as Fast Fourier Transform, Discrete Fourier Transform, Short-Time Fourier Transform, Principal Component Analysis, Wavelet Transform, Directed Transfer Function etc. have been used to analyze the recorded signals so as to unravel the distinctive event-related potential patterns in individuals with ASD. The review concludes that ERP proves to be an efficient tool in detecting the brain abnormalities and connectivity issues, indicating the heterogeneity of ASD. Many advanced techniques are utilized to decipher the underlying neural circuitry so as to aid in therapeutic interventions for improving the core areas of deficits.

EEG-Annotate: Automated identification and labeling of events in continuous signals with applications to EEG

BACKGROUND

In controlled laboratory EEG experiments, researchers carefully mark events and analyze subject responses time-locked to these events. Unfortunately, such markers may not be available or may come with poor timing resolution for experiments conducted in less-controlled naturalistic environments.

NEW METHOD

We present an integrated event-identification method for identifying particular responses that occur in unlabeled continuously recorded EEG signals based on information from recordings of other subjects potentially performing related tasks. We introduce the idea of timing slack and timing-tolerant performance measures to deal with jitter inherent in such non-time-locked systems. We have developed an implementation available as an open-source MATLAB toolbox (http://github.com/VisLab/EEG-Annotate) and have made test data available in a separate data note.

RESULTS

We applied the method to identify visual presentation events (both target and non-target) in data from an unlabeled subject using labeled data from other subjects with good sensitivity and specificity. The method also identified actual visual presentation events in the data that were not previously marked in the experiment.

COMPARISON WITH EXISTING METHODS

Although the method uses traditional classifiers for initial stages, the problem of identifying events based on the presence of stereotypical EEG responses is the converse of the traditional stimulus-response paradigm and has not been addressed in its current form.

CONCLUSIONS

In addition to identifying potential events in unlabeled or incompletely labeled EEG, these methods also allow researchers to investigate whether particular stereotypical neural responses are present in other circumstances. Timing-tolerance has the added benefit of accommodating inter- and intra- subject timing variations.

LATENT CLASS MODELING USING MATRIX COVARIATES WITH APPLICATION TO IDENTIFYING EARLY PLACEBO RESPONDERS BASED ON EEG SIGNALS

Latent class models are widely used to identify unobserved subgroups (i.e., latent classes) based upon one or more manifest variables. The probability of belonging to each subgroup is typically modeled as a function of a set of measured covariates. In this paper, we extend existing latent class models to incorporate matrix covariates. This research is motivated by a randomized placebo-controlled depression clinical trial. One study goal is to identify a subgroup of subjects who experience symptoms improvement early on during antidepressant treatment, which is considered to be an indication of a placebo rather than a true pharmacological response. We want to relate the likelihood of belonging to this subgroup of early responders to baseline electroencephalography (EEG) measurement that takes the form of a matrix. The proposed method is built upon a low rank Candecomp/Parafac (CP) decomposition of the target coefficient matrix through low-dimensional latent variables, which effectively reduces the model dimensionality. We adopt a Bayesian hierarchical modeling approach to estimate the latent variables, which allows a flexible way to incorporate prior knowledge about covariate effect heterogeneity and offers a data-driven method of regularization. Simulation studies suggest that the proposed method is robust against potentially misspecified rank in the CP decomposition. With the motivating example we show how the proposed method can be applied to extract valuable information from baseline EEG measurements that explains the likelihood of belonging to the early responder subgroup, helping to identify placebo responders and suggesting new targets for the study of placebo response.

A wavelet-based technique to predict treatment outcome for Major Depressive Disorder

Treatment management for Major Depressive Disorder (MDD) has been challenging. However, electroencephalogram (EEG)-based predictions of antidepressant’s treatment outcome may help during antidepressant’s selection and ultimately improve the quality of life for MDD patients. In this study, a machine learning (ML) method involving pretreatment EEG data was proposed to perform such predictions for Selective Serotonin Reuptake Inhibitor (SSRIs). For this purpose, the acquisition of experimental data involved 34 MDD patients and 30 healthy controls. Consequently, a feature matrix was constructed involving time-frequency decomposition of EEG data based on wavelet transform (WT) analysis, termed as EEG data matrix. However, the resultant EEG data matrix had high dimensionality. Therefore, dimension reduction was performed based on a rank-based feature selection method according to a criterion, i.e., receiver operating characteristic (ROC). As a result, the most significant features were identified and further be utilized during the training and testing of a classification model, i.e., the logistic regression (LR) classifier. Finally, the LR model was validated with 100 iterations of 10-fold cross-validation (10-CV). The classification results were compared with short-time Fourier transform (STFT) analysis, and empirical mode decompositions (EMD). The wavelet features extracted from frontal and temporal EEG data were found statistically significant. In comparison with other time-frequency approaches such as the STFT and EMD, the WT analysis has shown highest classification accuracy, i.e., accuracy = 87.5%, sensitivity = 95%, and specificity = 80%. In conclusion, significant wavelet coefficients extracted from frontal and temporal pre-treatment EEG data involving delta and theta frequency bands may predict antidepressant’s treatment outcome for the MDD patients.