Abnormal early dynamic individual patterns of functional networks in low gamma band for depression recognition

Sampling inequalities affect generalization of neuroimaging-based diagnostic classifiers in psychiatry

BACKGROUND

The development of machine learning models for aiding in the diagnosis of mental disorder is recognized as a significant breakthrough in the field of psychiatry. However, clinical practice of such models remains a challenge, with poor generalizability being a major limitation.

METHODS

Here, we conducted a pre-registered meta-research assessment on neuroimaging-based models in the psychiatric literature, quantitatively examining global and regional sampling issues over recent decades, from a view that has been relatively underexplored. A total of 476 studies (n = 118,137) were included in the current assessment. Based on these findings, we built a comprehensive 5-star rating system to quantitatively evaluate the quality of existing machine learning models for psychiatric diagnoses.

RESULTS

A global sampling inequality in these models was revealed quantitatively (sampling Gini coefficient (G) = 0.81, p < .01), varying across different countries (regions) (e.g., China, G = 0.47; the USA, G = 0.58; Germany, G = 0.78; the UK, G = 0.87). Furthermore, the severity of this sampling inequality was significantly predicted by national economic levels (β =  - 2.75, p < .001, R²_adj = 0.40; r =  - .84, 95% CI: - .41 to - .97), and was plausibly predictable for model performance, with higher sampling inequality for reporting higher classification accuracy. Further analyses showed that lack of independent testing (84.24% of models, 95% CI: 81.0-87.5%), improper cross-validation (51.68% of models, 95% CI: 47.2-56.2%), and poor technical transparency (87.8% of models, 95% CI: 84.9-90.8%)/availability (80.88% of models, 95% CI: 77.3-84.4%) are prevailing in current diagnostic classifiers despite improvements over time. Relating to these observations, model performances were found decreased in studies with independent cross-country sampling validations (all p < .001, BF₁₀ > 15). In light of this, we proposed a purpose-built quantitative assessment checklist, which demonstrated that the overall ratings of these models increased by publication year but were negatively associated with model performance.

CONCLUSIONS

Together, improving sampling economic equality and hence the quality of machine learning models may be a crucial facet to plausibly translating neuroimaging-based diagnostic classifiers into clinical practice.

Evaluation of Risk of Bias in Neuroimaging-Based Artificial Intelligence Models for Psychiatric Diagnosis: A Systematic Review

IMPORTANCE

Neuroimaging-based artificial intelligence (AI) diagnostic models have proliferated in psychiatry. However, their clinical applicability and reporting quality (ie, feasibility) for clinical practice have not been systematically evaluated.

OBJECTIVE

To systematically assess the risk of bias (ROB) and reporting quality of neuroimaging-based AI models for psychiatric diagnosis.

EVIDENCE REVIEW

PubMed was searched for peer-reviewed, full-length articles published between January 1, 1990, and March 16, 2022. Studies aimed at developing or validating neuroimaging-based AI models for clinical diagnosis of psychiatric disorders were included. Reference lists were further searched for suitable original studies. Data extraction followed the CHARMS (Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modeling Studies) and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines. A closed-loop cross-sequential design was used for quality control. The PROBAST (Prediction Model Risk of Bias Assessment Tool) and modified CLEAR (Checklist for Evaluation of Image-Based Artificial Intelligence Reports) benchmarks were used to systematically evaluate ROB and reporting quality.

FINDINGS

A total of 517 studies presenting 555 AI models were included and evaluated. Of these models, 461 (83.1%; 95% CI, 80.0%-86.2%) were rated as having a high overall ROB based on the PROBAST. The ROB was particular high in the analysis domain, including inadequate sample size (398 of 555 models [71.7%; 95% CI, 68.0%-75.6%]), poor model performance examination (with 100% of models lacking calibration examination), and lack of handling data complexity (550 of 555 models [99.1%; 95% CI, 98.3%-99.9%]). None of the AI models was perceived to be applicable to clinical practices. Overall reporting completeness (ie, number of reported items/number of total items) for the AI models was 61.2% (95% CI, 60.6%-61.8%), and the completeness was poorest for the technical assessment domain with 39.9% (95% CI, 38.8%-41.1%).

CONCLUSIONS AND RELEVANCE

This systematic review found that the clinical applicability and feasibility of neuroimaging-based AI models for psychiatric diagnosis were challenged by a high ROB and poor reporting quality. Particularly in the analysis domain, ROB in AI diagnostic models should be addressed before clinical application.

Gamma band VMPFC-PreCG.L connection variation after the onset of negative emotional stimuli can predict mania in depressive patients

OBJECTIVE

Because of the similar clinical symptoms, it is difficult to distinguish unipolar disorder (UD) from bipolar disorder (BD) in the depressive episode using the available clinical features, especially for those who meet the diagnostic criteria of UD, however, experience the manic episode during the follow-up (tBD).

METHODS

Magnetoencephalography recordings during a sad expression recognition task were obtained from 81 patients (27 BD, 24 tBD, 30 UD) and 26 healthy controls (HCs). Source analysis was applied to localize 64 regions of interest in the low gamma band (30-50 Hz). Regional functional connections (FCs) were constructed respectively within three time periods (early: 0-200 ms, middle: 200-400 ms, and post: 400-600 ms). The network-based statistic method was used to explore the abnormal connection patterns in tBD compared to UD and HC. BD was applied to explore whether such abnormality is still significant between every two groups of BD, tBD, UD, and HC.

RESULTS

The VMPFC-PreCG.L connection was found to be a significantly different connection between tBD and UD in the early time period and between tBD and BD in the middle time period. Furthermore, the middle/early time period ratio of FC value of VMPFC-PreCG.L connection was negatively correlated with the bipolarity index in tBD.

CONCLUSIONS

The VMPFC-PreCG.L connection in different time periods after the onset of sad facial stimuli may be a potential biomarker to distinguish the different states of BD. The FC ratio of VMPFC-PreCG.L connection may predict whether patients with depressive episodes subsequently develop mania.

An insight into diagnosis of depression using machine learning techniques: a systematic review

BACKGROUND

In this modern era, depression is one of the most prevalent mental disorders from which millions of individuals are affected today. The symptoms of depression are heterogeneous and often coincide with other disorders such as bipolar disorder, Parkinson's, schizophrenia, etc. It is a serious mental illness that may lead to other health problems if left untreated. Currently, identifying individuals with depression is totally based on the expertise of the clinician's experience. In order to assist clinicians in identifying the characteristics and classifying depressed people, different types of data modalities and machine learning techniques have been incorporated by researchers in this field. This study aims to find the answers to some important questions related to the trend of publications, data modality, machine learning models, dataset usage, pre-processing techniques and feature extraction and selection techniques that are prevalent and guide the direction of future research on depression diagnosis.

METHODS

This systematic review was conducted using a broad range of articles from two major databases: IEEE Xplore and PubMed. Studies ranging from the years 2011 to April 2021 were retrieved from the databases resulting in a total of 590 articles (53 articles from the IEEE Xplore database and 537 articles from the PubMed database). Out of those, the articles which satisfied the defined inclusion criteria were investigated for further analysis.

RESULTS

A total of 135 articles were identified and analysed for this review. High growth in the number of publications has been observed in recent years. Furthermore, significant diversity in the use of data modalities and machine learning classifiers has also been noted in this study. fMRI data with an SVM classifier was found to be the most popular choice among researchers. In most of the studies, data scarcity and small sample size, particularly for neuroimaging data are major concerns. The use of identical data pre-processing tools for similar data modalities can be seen. This study also provides statistical analysis of the current framework with respect to the modality, machine learning classifier, sample size and accuracy by applying one-way ANOVA and the Tukey - Kramer test.

CONCLUSION

The results indicate that an effective fusion of machine learning techniques with a potential data modality has a promising future for assisting clinicians in automatic depression diagnosis.

An electroencephalography connectome predictive model of major depressive disorder severity

Emerging evidence showed that major depressive disorder (MDD) is associated with disruptions of brain structural and functional networks, rather than impairment of isolated brain region. Thus, connectome-based models capable of predicting the depression severity at the individual level can be clinically useful. Here, we applied a machine-learning approach to predict the severity of depression using resting-state networks derived from source-reconstructed Electroencephalography (EEG) signals. Using regression models and three independent EEG datasets (N = 328), we tested whether resting state functional connectivity could predict individual depression score. On the first dataset, results showed that individuals scores could be reasonably predicted (r = 0.6, p = 4 × 10-18) using intrinsic functional connectivity in the EEG alpha band (8-13 Hz). In particular, the brain regions which contributed the most to the predictive network belong to the default mode network. We further tested the predictive potential of the established model by conducting two external validations on (N1 = 53, N2 = 154). Results showed statistically significant correlations between the predicted and the measured depression scale scores (r1 = 0.52, r2 = 0.44, p < 0.001). These findings lay the foundation for developing a generalizable and scientifically interpretable EEG network-based markers that can ultimately support clinicians in a biologically-based characterization of MDD.

Electrophysiological correlates and predictors of the antidepressant response to repeated ketamine infusions in treatment-resistant depression

BACKGROUND

Sub-anesthetic ketamine doses rapidly reduce depressive symptoms, although additional investigations of the underlying neural mechanisms and the prediction of response outcomes are needed. Electroencephalographic (EEG)-derived measures have shown promise in predicting antidepressant response to a variety of treatments, and are sensitive to ketamine administration. This study examined their utility in characterizing changes in depressive symptoms following single and repeated ketamine infusions.

METHODS

Recordings were obtained from patients with treatment-resistant major depressive disorder (MDD) (N = 24) enrolled in a multi-phase clinical ketamine trial. During the randomized, double-blind, crossover phase (Phase 1), patients received intravenous ketamine (0.5 mg/kg) and midazolam (30 μg/kg), at least 1 week apart. For each medication, three resting, eyes-closed recordings were obtained per session (pre-infusion, immediately post-infusion, 2 h post-infusion), and changes in power (delta, theta1/2/total, alpha1/2/total, beta, gamma), alpha asymmetry, theta cordance, and theta source-localized anterior cingulate cortex activity were quantified. The relationships between ketamine-induced changes with early (Phase 1) and sustained (Phases 2,3: open-label repeated infusions) decreases in depressive symptoms (Montgomery-Åsberg Depression Rating Score, MADRS) and suicidal ideation (MADRS item 10) were examined.

RESULTS

Both medications decreased alpha and theta immediately post-infusion, however, only midazolam increased delta (post-infusion), and only ketamine increased gamma (immediately post- and 2 h post-infusion). Regional- and frequency-specific ketamine-induced EEG changes were related to and predictive of decreases in depressive symptoms (theta, gamma) and suicidal ideation (alpha). Early and sustained treatment responders differed at baseline in surface-level and source-localized theta.

CONCLUSIONS

Ketamine exerts frequency-specific changes on EEG-derived measures, which are related to depressive symptom decreases in treatment-resistant MDD and provide information regarding early and sustained individual response to ketamine.

CLINICAL TRIAL REGISTRATION

ClinicalTrials.gov: Action of Ketamine in Treatment-Resistant Depression, NCT01945047.

Investigating the Effects of Auditory and Vibrotactile Rhythmic Sensory Stimulation on Depression: An EEG Pilot Study

Background

Major depressive disorder (MDD) is a persistent psychiatric condition and one of the leading causes of global disease burden. In a previous study, we investigated the effects of a five-week intervention consisting of rhythmic gamma frequency (30-70 Hz) vibroacoustic stimulation in 20 patients formally diagnosed with MDD. In that study, the findings suggested a significant clinical improvement in depression symptoms as measured using the Montgomery-Asberg Depression Rating Scale (MADRS), with 37% of participants meeting the criteria for clinical response. The goal of the present research was to examine possible changes from baseline to posttreatment in resting-state electroencephalography (EEG) recordings using the same treatment protocol and to characterize basic changes in EEG related to treatment response.

Materials and methods

The study sample consisted of 19 individuals aged 18-70 years with a clinical diagnosis of MDD. The participants were assessed before and after a five-week treatment period, which consisted of listening to an instrumental musical track on a vibroacoustic device, delivering auditory and vibrotactile stimulus in the gamma-band range (30-70 Hz, with particular emphasis on 40 Hz). The primary outcome measure was the change in Montgomery-Asberg Depression Rating Scale (MADRS) from baseline to posttreatment and resting-state EEG.

Results

Analysis comparing MADRS score at baseline and post-intervention indicated a significant change in the severity of depression symptoms after five weeks (t = 3.9923, df = 18, p = 0.0009). The clinical response rate was 36.85%. Resting-state EEG power analysis revealed a significant increase in occipital alpha power (t = -2.149, df = 18, p = 0.04548), as well as an increase in the prefrontal gamma power of the responders (t = 2.8079, df = 13.431, p = 0.01442).

Conclusions

The results indicate that improvements in MADRS scores after rhythmic sensory stimulation (RSS) were accompanied by an increase in alpha power in the occipital region and an increase in gamma in the prefrontal region, thus suggesting treatment effects on cortical activity in depression. The results of this pilot study will help inform subsequent controlled studies evaluating whether treatment response to vibroacoustic stimulation constitutes a real and replicable reduction of depressive symptoms and to characterize the underlying mechanisms.

Depression Diagnosis Modeling With Advanced Computational Methods: Frequency-Domain eMVAR and Deep Learning

Electroencephalogram (EEG)-based automated depression diagnosis systems have been suggested for early and accurate detection of mood disorders. EEG signals are highly irregular, nonlinear, and nonstationary in nature and are traditionally studied from a linear viewpoint by means of statistical and frequency features. Since, linear metrics present certain limitations and nonlinear methods have proven to be an efficient tool in understanding the complexities of the brain in the identification of underlying behavior of biological signals, such as electrocardiogram, EEG and magnetoencephalogram and thus, can be applied to all nonstationary signals. Various nonlinear algorithms can be used in the analysis of EEG signals. In this research paper, we aim to develop a novel methodology for EEG-based depression diagnosis utilizing 2 advanced computational techniques: frequency-domain extended multivariate autoregressive (eMVAR) and deep learning (DL). We proposed a hybrid method comprising a pretrained ResNet-50 and long-short term memory (LSTM) to capture depression-specific information and compared with a strong conventional machine learning (ML) framework having eMVAR connectivity features. The following 8 causality measures, which interpret the interaction mechanisms among spectrally decomposed oscillations, were used to extract features from multivariate EEG time series: directed coherence (DC), directed transfer function (DTF), partial DC (PDC), generalized PDC (gPDC), extended DC (eDC), delayed DC (dDC), extended PDC (ePDC), and delayed PDC (dPDC). The classification accuracies were 84% with DC, 85% with DTF, 95.3% with PDC, 95.1% with gPDC, 84.8% with eDC, 84.6% with dDC, 84.2% with ePDC, and 95.9% with dPDC for the eMVAR framework. Through a DL framework (ResNet-50 + LSTM), the classification accuracy was achieved as 90.22%. The results demonstrate that our DL methodology is a competitive alternative to the strong feature extraction-based ML methods in depression classification.

The Benefits of Music Listening for Induced State Anxiety: Behavioral and Physiological Evidence

BACKGROUND

Some clinical studies have indicated that neutral and happy music may relieve state anxiety. However, the brain mechanisms by which these effective interventions in music impact state anxiety remain unknown.

METHODS

In this study, we selected music with clinical effects for therapy, and 62 subjects were included using the evoked anxiety paradigm. After evoking anxiety with a visual stimulus, all subjects were randomly divided into three groups (listening to happy music, neutral music and a blank stimulus), and EEG signals were acquired.

RESULTS

We found that different emotional types of music might have different mechanisms in state anxiety interventions. Neutral music had the effect of alleviating state anxiety. The brain mechanisms supported that neutral music ameliorating state anxiety was associated with decreased power spectral density of the occipital lobe and increased brain functional connectivity between the occipital lobe and frontal lobe. Happy music also had the effect of alleviating state anxiety, and the brain mechanism was associated with enhanced brain functional connectivity between the occipital lobe and right temporal lobe.

CONCLUSIONS

This study may be important for a deep understanding of the mechanisms associated with state anxiety music interventions and may further contribute to future clinical treatment using nonpharmaceutical interventions.

Positive AMPA receptor modulation in the treatment of neuropsychiatric disorders: A long and winding road

Glutamatergic transmission is widely implicated in neuropsychiatric disorders, and the discovery that ketamine elicits rapid-acting antidepressant effects by modulating α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor (AMPAR) signaling has spurred a resurgence of interest in the field. This review explores agents in various stages of development for neuropsychiatric disorders that positively modulate AMPARs, both directly and indirectly. Despite promising preclinical research, few direct and indirect AMPAR positive modulators have progressed past early clinical development. Challenges such as low potency have created barriers to effective implementation. Nevertheless, the functional complexity of AMPARs sets them apart from other drug targets and allows for specificity in drug discovery. Additional effective treatments for neuropsychiatric disorders that work through positive AMPAR modulation may eventually be developed.

The Oscillatory Profile Induced by the Anxiogenic Drug FG-7142 in the Amygdala-Hippocampal Network Is Reversed by Infralimbic Deep Brain Stimulation: Relevance for Mood Disorders

Anxiety and depression exhibit high comorbidity and share the alteration of the amygdala–hippocampal–prefrontal network, playing different roles in the ventral and dorsal hippocampi. Deep brain stimulation of the infralimbic cortex in rodents or the human equivalent—the subgenual cingulate cortex—constitutes a fast antidepressant treatment. The aim of this work was: (1) to describe the oscillatory profile in a rodent model of anxiety, and (2) to deepen the therapeutic basis of infralimbic deep brain stimulation in mood disorders. First, the anxiogenic drug FG-7142 was administered to anaesthetized rats to characterize neural oscillations within the amygdala and the dorsoventral axis of the hippocampus. Next, deep brain stimulation was applied. FG-7142 administration drastically reduced the slow waves, increasing delta, low theta, and beta oscillations in the network. Moreover, FG-7142 altered communication in these bands in selective subnetworks. Deep brain stimulation of the infralimbic cortex reversed most of these FG-7142 effects. Cross-frequency coupling was also inversely modified by FG-7142 and by deep brain stimulation. Our study demonstrates that the hyperactivated amygdala–hippocampal network associated with the anxiogenic drug exhibits an oscillatory fingerprint. The study contributes to comprehending the neurobiological basis of anxiety and the effects of infralimbic deep brain stimulation.

Maternal separation induces changes in TREK-1 and 5HT1A expression in brain areas involved in the stress response in a sex-dependent way

Depression is a prevalent disease in modern society, and has been linked to stressful events at early ages. Women are more susceptible to depression, and the neural basis for this are still under investigation. Serotonin is known to be involved in depression, and a decrease in 5HT_1A expression is observed on temporal and cortical areas in both men and women with depression. As knockout animals for TREK-1 are resilient to depression, this channel has emerged as a new potential pharmacological target for depression treatment. In this study, maternal separation (MS) was used to emulate early-life stress, and evaluate behaviour, as well as TREK-1 and 5HT_1A expression in the brain using immunohistochemistry. In juvenile females, 5HT_1A reduction coupled to increased TREK-1 in the dentate gyrus (DG) was associated with behavioural despair, as well as increased TREK-1 expression in basolateral amygdala (BLA) and prelimbic cortex (PL). In juvenile males, MS induced an increase in 5HT_1A in the BLA, and in TREK-1 in the PL, while no behavioural despair was observed. Anhedonia and anxiety-like behaviour were not induced by MS. We conclude stress-induced increase in TREK-1 in PL and GD is associated to depression, while 5HT_1A changes coupled to TREK-1 changes may be necessary to induce depression, with females being more vulnerable to MS effects than males. Thus, TREK-1 and 5HT_1A may be potential pharmacological targets for antidepressants development.

Electrophysiological biomarkers of antidepressant response to ketamine in treatment-resistant depression: Gamma power and long-term potentiation

Over the last two decades, the discovery of ketamine's antidepressant properties has galvanized research into the neurobiology of treatment-resistant depression. Nevertheless, the mechanism of action underlying antidepressant response to ketamine remains unclear. This study reviews electrophysiological studies of ketamine's effects in individuals with depression as well as healthy controls, with a focus on two putative markers of synaptic potentiation: gamma oscillations and long-term potentiation. The review focuses on: 1) measures of gamma oscillations and power and their relationship to both acute, psychotomimetic drug effects as well as delayed antidepressant response in mood disorders; 2) changes in long-term potentiation as a promising measure of synaptic potentiation following ketamine administration; and 3) recent efforts to model antidepressant response to ketamine using novel computational modeling techniques, in particular the application of dynamic causal modeling to electrophysiological data. The latter promises to better characterize the mechanisms underlying ketamine's antidepressant effects.

An enriched granger causal model allowing variable static anatomical constraints

The anatomical connectivity constrains but does not fully determine functional connectivity, especially when one explores into the dynamics over the course of a trial. Therefore, an enriched granger causal model (GCM) integrated with anatomical prior information is proposed in this study, to describe the dynamic effective connectivity to distinguish the depression and explore the pathogenesis of depression. In the proposed frame, the anatomical information was converted via an optimized transformation model, which was then integrated into the normal GCM by variational bayesian model. Magnetoencephalography (MEG) signals and diffusion tensor imaging (DTI) of 24 depressive patients and 24 matched controls were utilized for performance comparison. Together with the sliding windowed MEG signals under sad facial stimuli, the enriched GCM was applied to calculate the regional-pair dynamic effective connectivity, which were repeatedly sifted via feature selection and fed into different classifiers. From the aspects of model errors and recognition accuracy rates, results supported the superiority of the enriched GCM with anatomical priors over the normal GCM. For the effective connectivity with anatomical priors, the best subject discrimination accuracy of SVM was 85.42% (the sensitivity was 87.50% and the specificity was 83.33%). Furthermore, discriminative feature analysis suggested that the enriched GCM that detect the variable anatomical constraint on function could better detect more stringent and less dynamic brain function in depression. The proposed approach is valuable in dynamic functional dysfunction exploration in depression and could be useful for depression recognition.