A machine learning approach using EEG data to predict response to SSRI treatment for major depressive disorder

Prediction of treatment response in major depressive disorder using a hybrid of convolutional recurrent deep neural networks and effective connectivity based on EEG signal

In this study, we have developed a novel method based on deep learning and brain effective connectivity to classify responders and non-responders to selective serotonin reuptake inhibitors (SSRIs) antidepressants in major depressive disorder (MDD) patients prior to the treatment using EEG signal. The effective connectivity of 30 MDD patients was determined by analyzing their pretreatment EEG signals, which were then concatenated into delta, theta, alpha, and beta bands and transformed into images. Using these images, we then fine tuned a hybrid Convolutional Neural Network that is enhanced with bidirectional Long Short-Term Memory cells based on transfer learning. The Inception-v3, ResNet18, DenseNet121, and EfficientNet-B0 models are implemented as base models. Finally, the models are followed by BiLSTM and dense layers in order to classify responders and non-responders to SSRI treatment. Results showed that the EfficiencyNet-B0 has the highest accuracy of 98.33, followed by DensNet121, ResNet18 and Inception-v3. Therefore, a new method was proposed in this study that uses deep learning models to extract both spatial and temporal features automatically, which will improve classification results. The proposed method provides accurate identification of MDD patients who are responding, thereby reducing the cost of medical facilities and patient care.

Machine learning reveals differential effects of depression and anxiety on reward and punishment processing

Recent studies suggest that depression and anxiety are associated with unique aspects of EEG responses to reward and punishment, respectively; also, abnormal responses to punishment in depressed individuals are related to anxiety, the symptoms of which are comorbid with depression. In a non-clinical sample, we aimed to investigate the relationships between reward processing and anxiety, between punishment processing and anxiety, between reward processing and depression, and between punishment processing and depression. Towards this aim, we separated feedback-related brain activity into delta and theta bands to isolate activity that indexes functionally distinct processes. Based on the delta/theta frequency and feedback valence, we then used machine learning (ML) to classify individuals with high severity of depressive symptoms and individuals with high severity of anxiety symptoms versus controls. The significant difference between the depression and control groups was driven mainly by delta activity; there were no differences between reward- and punishment-theta activities. The high severity of anxiety symptoms was marginally more strongly associated with the punishment- than the reward-theta feedback processing. The findings provide new insights into the differences in the impacts of anxiety and depression on reward and punishment processing; our study shows the utility of ML in testing brain-behavior hypotheses and emphasizes the joint effect of theta-RewP/FRN and delta frequency on feedback-related brain activity.

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.

Benchmarking Mental Health Status Using Passive Sensor Data: Protocol for a Prospective Observational Study

BACKGROUND

Computational psychiatry has the potential to advance the diagnosis, mechanistic understanding, and treatment of mental health conditions. Promising results from clinical samples have led to calls to extend these methods to mental health risk assessment in the general public; however, data typically used with clinical samples are neither available nor scalable for research in the general population. Digital phenotyping addresses this by capitalizing on the multimodal and widely available data created by sensors embedded in personal digital devices (eg, smartphones) and is a promising approach to extending computational psychiatry methods to improve mental health risk assessment in the general population.

OBJECTIVE

Building on recommendations from existing computational psychiatry and digital phenotyping work, we aim to create the first computational psychiatry data set that is tailored to studying mental health risk in the general population; includes multimodal, sensor-based behavioral features; and is designed to be widely shared across academia, industry, and government using gold standard methods for privacy, confidentiality, and data integrity.

METHODS

We are using a stratified, random sampling design with 2 crossed factors (difficulties with emotion regulation and perceived life stress) to recruit a sample of 400 community-dwelling adults balanced across high- and low-risk for episodic mental health conditions. Participants first complete self-report questionnaires assessing current and lifetime psychiatric and medical diagnoses and treatment, and current psychosocial functioning. Participants then complete a 7-day in situ data collection phase that includes providing daily audio recordings, passive sensor data collected from smartphones, self-reports of daily mood and significant events, and a verbal description of the significant daily events during a nightly phone call. Participants complete the same baseline questionnaires 6 and 12 months after this phase. Self-report questionnaires will be scored using standard methods. Raw audio and passive sensor data will be processed to create a suite of daily summary features (eg, time spent at home).

RESULTS

Data collection began in June 2022 and is expected to conclude by July 2024. To date, 310 participants have consented to the study; 149 have completed the baseline questionnaire and 7-day intensive data collection phase; and 61 and 31 have completed the 6- and 12-month follow-up questionnaires, respectively. Once completed, the proposed data set will be made available to academic researchers, industry, and the government using a stepped approach to maximize data privacy.

CONCLUSIONS

This data set is designed as a complementary approach to current computational psychiatry and digital phenotyping research, with the goal of advancing mental health risk assessment within the general population. This data set aims to support the field's move away from siloed research laboratories collecting proprietary data and toward interdisciplinary collaborations that incorporate clinical, technical, and quantitative expertise at all stages of the research process.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

DERR1-10.2196/53857.

Resting-state EEG connectivity recorded before and after rTMS treatment in patients with treatment-resistant depression

Repetitive transcranial magnetic stimulation (rTMS) has shown efficacy and tolerability in Major Depressive Disorder (MDD). However, the underlying mechanisms of its antidepressant effects remain unclear. This open-label study investigated electroencephalography (EEG) functional connectivity markers associated with response and the antidepressant effects of rTMS. Resting-state EEG data were collected from 28 participants with MDD before and after a four-week rTMS course. Source-space functional connectivity between 38 cortical regions was compared using an orthogonalised amplitude approach. Depressive symptoms significantly improved following rTMS, with 43 % of participants classified as responders. While the study's functional connectivity findings did not withstand multiple comparison corrections, exploratory analyses suggest an association between theta band connectivity and rTMS treatment mechanisms. Fronto-parietal theta connectivity increased after treatment but did not correlate with antidepressant response. Notably, low baseline theta connectivity was associated with greater response. However, due to the exploratory nature and small sample size, further replication is needed. The findings provide preliminary evidence that EEG functional connectivity, particularly within the theta band, may reflect the mechanisms by which rTMS exerts its therapeutic effects.

Investigation of factors regarding the effects of COVID-19 pandemic on college students' depression by quantum annealer

Diverse cases regarding the impact, with its related factors, of the COVID-19 pandemic on mental health have been reported in previous studies. In this study, multivariable datasets were collected from 751 college students who could be easily affected by pandemics based on the complex relationships between various mental health factors. We utilized quantum annealing (QA)-based feature selection algorithms that were executed by commercial D-Wave quantum computers to determine the changes in the relative importance of the associated factors before and after the pandemic. Multivariable linear regression (MLR) and XGBoost models were also applied to validate the QA-based algorithms. Based on the experimental results, we confirm that QA-based algorithms have comparable capabilities in factor analysis research to the MLR models that have been widely used in previous studies. Furthermore, the performance of the QA-based algorithms was validated through the important factor results from the algorithms. Pandemic-related factors (e.g., confidence in the social system) and psychological factors (e.g. decision-making in uncertain situations) were more important in post-pandemic conditions. Although the results should be validated using other mental health variables or national datasets, this study will serve as a reference for researchers regarding the use of the quantum annealing approach in factor analysis with validation through real-world survey dataset analysis.

Developing an Electroencephalography-Based Model for Predicting Response to Antidepressant Medication

Importance

Untreated depression is a growing public health concern, with patients often facing a prolonged trial-and-error process in search of effective treatment. Developing a predictive model for treatment response in clinical practice remains challenging.

Objective

To establish a model based on electroencephalography (EEG) to predict response to 2 distinct selective serotonin reuptake inhibitor (SSRI) medications.

Design, Setting, and Participants

This prognostic study developed a predictive model using EEG data collected between 2011 and 2017 from 2 independent cohorts of participants with depression: 1 from the first Canadian Biomarker Integration Network in Depression (CAN-BIND) group and the other from the Establishing Moderators and Biosignatures of Antidepressant Response for Clinical Care (EMBARC) consortium. Eligible participants included those aged 18 to 65 years who had a diagnosis of major depressive disorder. Data were analyzed from January to December 2022.

Exposures

In an open-label trial, CAN-BIND participants received an 8-week treatment regimen of escitalopram treatment (10-20 mg), and EMBARC participants were randomized in a double-blind trial to receive an 8-week sertraline (50-200 mg) treatment or placebo treatment.

Main Outcomes and Measures

The model's performance was estimated using balanced accuracy, specificity, and sensitivity metrics. The model used data from the CAN-BIND cohort for internal validation, and data from the treatment group of the EMBARC cohort for external validation. At week 8, response to treatment was defined as a 50% or greater reduction in the primary, clinician-rated scale of depression severity.

Results

The CAN-BIND cohort included 125 participants (mean [SD] age, 36.4 [13.0] years; 78 [62.4%] women), and the EMBARC sertraline treatment group included 105 participants (mean [SD] age, 38.4 [13.8] years; 72 [68.6%] women). The model achieved a balanced accuracy of 64.2% (95% CI, 55.8%-72.6%), sensitivity of 66.1% (95% CI, 53.7%-78.5%), and specificity of 62.3% (95% CI, 50.1%-73.8%) during internal validation with CAN-BIND. During external validation with EMBARC, the model achieved a balanced accuracy of 63.7% (95% CI, 54.5%-72.8%), sensitivity of 58.8% (95% CI, 45.3%-72.3%), and specificity of 68.5% (95% CI, 56.1%-80.9%). Additionally, the balanced accuracy for the EMBARC placebo group (118 participants) was 48.7% (95% CI, 39.3%-58.0%), the sensitivity was 50.0% (95% CI, 35.2%-64.8%), and the specificity was 47.3% (95% CI, 35.9%-58.7%), suggesting the model's specificity in predicting SSRIs treatment response.

Conclusions and Relevance

In this prognostic study, an EEG-based model was developed and validated in 2 independent cohorts. The model showed promising accuracy in predicting treatment response to 2 distinct SSRIs, suggesting potential applications for personalized depression treatment.

Irremediability in psychiatric euthanasia: examining the objective standard

BACKGROUND

Irremediability is a key requirement for euthanasia and assisted suicide for psychiatric disorders (psychiatric EAS). Countries like the Netherlands and Belgium ask clinicians to assess irremediability in light of the patient's diagnosis and prognosis and 'according to current medical understanding'. Clarifying the relevance of a default objective standard for irremediability when applied to psychiatric EAS is crucial for solid policymaking. Yet so far, a thorough examination of this standard is lacking.

METHODS

Using treatment-resistant depression (TRD) as a test case, through a scoping review in PubMed, we analyzed the state-of-the-art evidence for whether clinicians can accurately predict individual long-term outcome and single out irremediable cases, by examining the following questions: (1) What is the definition of TRD; (2) What are group-level long-term outcomes of TRD; and (3) Can clinicians make accurate individual outcome predictions in TRD?

RESULTS

A uniform definition of TRD is lacking, with over 150 existing definitions, mostly focused on psychopharmacological research. Available yet limited studies about long-term outcomes indicate that a majority of patients with long-term TRD show significant improvement over time. Finally, evidence about individual predictions in TRD using precision medicine is growing, but methodological shortcomings and varying predictive accuracies pose important challenges for its implementation in clinical practice.

CONCLUSION

Our findings support the claim that, as per available evidence, clinicians cannot accurately predict long-term chances of recovery in a particular patient with TRD. This means that the objective standard for irremediability cannot be met, with implications for policy and practice of psychiatric EAS.

Sub-Scalp Implantable Telemetric EEG (SITE) for the Management of Neurological and Behavioral Disorders beyond Epilepsy

Sub-scalp Implantable Telemetric EEG (SITE) devices are under development for the treatment of epilepsy. However, beyond epilepsy, continuous EEG analysis could revolutionize the management of patients suffering from all types of brain disorders. This article reviews decades of foundational EEG research, collected from short-term routine EEG studies of common neurological and behavioral disorders, that may guide future SITE management and research. Established quantitative EEG methods, like spectral EEG power density calculation combined with state-of-the-art machine learning techniques applied to SITE data, can identify new EEG biomarkers of neurological disease. From distinguishing syncopal events from seizures to predicting the risk of dementia, SITE-derived EEG biomarkers can provide clinicians with real-time information about diagnosis, treatment response, and disease progression.

Accurately predicting hit songs using neurophysiology and machine learning

Identifying hit songs is notoriously difficult. Traditionally, song elements have been measured from large databases to identify the lyrical aspects of hits. We took a different methodological approach, measuring neurophysiologic responses to a set of songs provided by a streaming music service that identified hits and flops. We compared several statistical approaches to examine the predictive accuracy of each technique. A linear statistical model using two neural measures identified hits with 69% accuracy. Then, we created a synthetic set data and applied ensemble machine learning to capture inherent non-linearities in neural data. This model classified hit songs with 97% accuracy. Applying machine learning to the neural response to 1st min of songs accurately classified hits 82% of the time showing that the brain rapidly identifies hit music. Our results demonstrate that applying machine learning to neural data can substantially increase classification accuracy for difficult to predict market outcomes.

Deep learning and machine learning in psychiatry: a survey of current progress in depression detection, diagnosis and treatment

Informatics paradigms for brain and mental health research have seen significant advances in recent years. These developments can largely be attributed to the emergence of new technologies such as machine learning, deep learning, and artificial intelligence. Data-driven methods have the potential to support mental health care by providing more precise and personalised approaches to detection, diagnosis, and treatment of depression. In particular, precision psychiatry is an emerging field that utilises advanced computational techniques to achieve a more individualised approach to mental health care. This survey provides an overview of the ways in which artificial intelligence is currently being used to support precision psychiatry. Advanced algorithms are being used to support all phases of the treatment cycle. These systems have the potential to identify individuals suffering from mental health conditions, allowing them to receive the care they need and tailor treatments to individual patients who are mostly to benefit. Additionally, unsupervised learning techniques are breaking down existing discrete diagnostic categories and highlighting the vast disease heterogeneity observed within depression diagnoses. Artificial intelligence also provides the opportunity to shift towards evidence-based treatment prescription, moving away from existing methods based on group averages. However, our analysis suggests there are several limitations currently inhibiting the progress of data-driven paradigms in care. Significantly, none of the surveyed articles demonstrate empirically improved patient outcomes over existing methods. Furthermore, greater consideration needs to be given to uncertainty quantification, model validation, constructing interdisciplinary teams of researchers, improved access to diverse data and standardised definitions within the field. Empirical validation of computer algorithms via randomised control trials which demonstrate measurable improvement to patient outcomes are the next step in progressing models to clinical implementation.

Machine learning approaches and non-linear processing of extracted components in frontal region to predict rTMS treatment response in major depressive disorder

Predicting the therapeutic result of repetitive transcranial magnetic stimulation (rTMS) treatment could save time and costs as ineffective treatment can be avoided. To this end, we presented a machine-learning-based strategy for classifying patients with major depression disorder (MDD) into responders (R) and nonresponders (NR) to rTMS treatment. Resting state EEG data were recorded using 32 electrodes from 88 MDD patients before treatment. Then, patients underwent 7 weeks of rTMS, and 46 of them responded to treatment. By applying Independent Component Analysis (ICA) on EEG, we identified the relevant brain sources as possible indicators of neural activity in the dorsolateral prefrontal cortex (DLPFC). This was served through estimating the generators of activity in the sensor domain. Subsequently, we added physiological information and placed certain terms and conditions to offer a far more realistic estimation than the classic EEG. Ultimately, those components mapped in accordance with the region of the DLPFC in the sensor domain were chosen. Features extracted from the relevant ICs time series included permutation entropy (PE), fractal dimension (FD), Lempel-Ziv Complexity (LZC), power spectral density, correlation dimension (CD), features based on bispectrum, frontal and prefrontal cordance, and a combination of them. The most relevant features were selected by a Genetic Algorithm (GA). For classifying two groups of R and NR, K-Nearest Neighbor (KNN), Support Vector Machine (SVM), and Multilayer Perceptron (MLP) were applied to predict rTMS treatment response. To evaluate the performance of classifiers, a 10-fold cross-validation method was employed. A statistical test was used to assess the capability of features in differentiating R and NR for further research. EEG characteristics that can predict rTMS treatment response were discovered. The strongest discriminative indicators were EEG beta power, the sum of bispectrum diagonal elements in delta and beta bands, and CD. The Combined feature vector classified R and NR with a high performance of 94.31% accuracy, 92.85% specificity, 95.65% sensitivity, and 92.85% precision using SVM. This result indicates that our proposed method with power and nonlinear and bispectral features from relevant ICs time-series can predict the treatment outcome of rTMS for MDD patients only by one session pretreatment EEG recording. The obtained results show that the proposed method outperforms previous methods.

Understanding mental health through computers: An introduction to computational psychiatry

Computational psychiatry recently established itself as a new tool in the study of mental disorders and problems. Integration of different levels of analysis is creating computational phenotypes with clinical and research values, and constructing a way to arrive at precision psychiatry are part of this new branch. It conceptualizes the brain as a computational organ that receives from the environment parameters to respond to challenges through calculations and algorithms in continuous feedback and feedforward loops with a permanent degree of uncertainty. Through this conception, one can seize an understanding of the cerebral and mental processes in the form of theories or hypotheses based on data. Using these approximations, a better understanding of the disorder and its different determinant factors facilitates the diagnostics and treatment by having an individual, ecologic, and holistic approach. It is a tool that can be used to homologate and integrate multiple sources of information given by several theoretical models. In conclusion, it helps psychiatry achieve precision and reproducibility, which can help the mental health field achieve significant advancement. This article is a narrative review of the basis of the functioning of computational psychiatry with a critical analysis of its concepts.

Neuroelectric Correlates of Human Sexuality: A Review and Meta-Analysis

Many reviews on sexual arousal in humans focus on different brain imaging methods and behavioral observations. Although neurotransmission in the brain is mainly performed through electrochemical signals, there are no systematic reviews of the electrophysiological correlates of sexual arousal. We performed a systematic search on this subject and reviewed 255 studies including various electrophysiological methods. Our results show how neuroelectric signals have been used to investigate genital somatotopy as well as basic genital physiology during sexual arousal and how cortical electric signals have been recorded during orgasm. Moreover, experiments on the interactions of cognition and sexual arousal in healthy subjects and in individuals with abnormal sexual preferences were analyzed as well as case studies on sexual disturbances associated with diseases of the nervous system. In addition, 25 studies focusing on brain potentials during the interaction of cognition and sexual arousal were eligible for meta-analysis. The results showed significant effect sizes for specific brain potentials during sexual stimulation (P3: Cohen's d = 1.82, N = 300, LPP: Cohen's d = 2.30, N = 510) with high heterogeneity between the combined studies. Taken together, our review shows how neuroelectric methods can consistently differentiate sexual arousal from other emotional states.

Electroencephalography for predicting antidepressant treatment success: A systematic review and meta-analysis

BACKGROUND

Patients suffering from major depressive disorder (MDD) regularly experience non-response to treatment for their depressive episode. Personalized clinical decision making could shorten depressive episodes and reduce patient suffering. Although no clinical tools are currently available, machine learning analysis of electroencephalography (EEG) shows promise in treatment response prediction.

METHODS

With a systematic review and meta-analysis, we evaluated the accuracy of EEG for individual patient response prediction. Importantly, we included only prediction studies that used cross-validation. We used a bivariate model to calculate prediction success, as expressed by area-under the curve, sensitivity and specificity. Furthermore, we analyzed prediction success for separate antidepressant interventions.

RESULTS

15 studies with 12 individual patient samples and a total of 479 patients were included. Research methods varied considerably between studies. Meta-analysis of results from this heterogeneous set of studies resulted in an area under the curve of 0.91, a sensitivity of 83 % (95 % CI 74-89 %), and a specificity of 86 % (95 % CI 81-90 %). Classification performance did not significantly differ between treatments. Although studies were all internally validated, no externally validated studies have been reported. We found substantial risk of bias caused by methodological shortcomings such as non-independent feature selection, though performance of non-biased studies was comparable.

LIMITATIONS

Sample sizes were relatively small and no study used external validation, increasing the risk of overestimation of accuracy.

CONCLUSIONS

Electroencephalography can predict the response to antidepressant treatment with high accuracy. However, future studies with more rigorous validation are needed to produce a clinical tool to guide interventions in MDD.

PROSPERO REGISTRATION NUMBER

CRD42021268169.

Gradient boosting decision-tree-based algorithm with neuroimaging for personalized treatment in depression

Introduction

Pretreatment positron emission tomography (PET) with 2-deoxy-2-[18F]fluoro-D-glucose (FDG) and magnetic resonance spectroscopy (MRS) may identify biomarkers for predicting remission (absence of depression). Yet, no such image-based biomarkers have achieved clinical validity. The purpose of this study was to identify biomarkers of remission using machine learning (ML) with pretreatment FDG-PET/MRS neuroimaging, to reduce patient suffering and economic burden from ineffective trials.

Methods

This study used simultaneous PET/MRS neuroimaging from a double-blind, placebo-controlled, randomized antidepressant trial on 60 participants with major depressive disorder (MDD) before initiating treatment. After eight weeks of treatment, those with ≤ 7 on 17-item Hamilton Depression Rating Scale were designated a priori as remitters (free of depression, 37%). Metabolic rate of glucose uptake (metabolism) from 22 brain regions were acquired from PET. Concentrations (mM) of glutamine and glutamate and gamma-aminobutyric acid (GABA) in anterior cingulate cortex were quantified from MRS. The data were randomly split into 67% train and cross-validation (n = 40), and 33% test (n = 20) sets. The imaging features, along with age, sex, handedness, and treatment assignment (selective serotonin reuptake inhibitor or SSRI vs. placebo) were entered into the eXtreme Gradient Boosting (XGBoost) classifier for training.

Results

In test data, the model showed 62% sensitivity, 92% specificity, and 77% weighted accuracy. Pretreatment metabolism of left hippocampus from PET was the most predictive of remission.

Conclusions

The pretreatment neuroimaging takes around 60 minutes but has potential to prevent weeks of failed treatment trials. This study effectively addresses common issues for neuroimaging analysis, such as small sample size, high dimensionality, and class imbalance.

Toward biophysical markers of depression vulnerability

A major difficulty with treating psychiatric disorders is their heterogeneity: different neural causes can lead to the same phenotype. To address this, we propose describing the underlying pathophysiology in terms of interpretable, biophysical parameters of a neural model derived from the electroencephalogram. We analyzed data from a small patient cohort of patients with depression and controls. Using DCM, we constructed biophysical models that describe neural dynamics in a cortical network activated during a task that is used to assess depression state. We show that biophysical model parameters are biomarkers, that is, variables that allow subtyping of depression at a biological level. They yield a low dimensional, interpretable feature space that allowed description of differences between individual patients with depressive symptoms. They could capture internal heterogeneity/variance of depression state and achieve significantly better classification than commonly used EEG features. Our work is a proof of concept that a combination of biophysical models and machine learning may outperform earlier approaches based on classical statistics and raw brain data.

Computational approaches to treatment response prediction in major depression using brain activity and behavioral data: A systematic review

Major depressive disorder is a heterogeneous diagnostic category with multiple available treatments. With the goal of optimizing treatment selection, researchers are developing computational models that attempt to predict treatment response based on various pretreatment measures. In this paper, we review studies that use brain activity data to predict treatment response. Our aim is to highlight and clarify important methodological differences between various studies that relate to the incorporation of domain knowledge, specifically within two approaches delineated as data-driven and theory-driven. We argue that theory-driven generative modeling, which explicitly models information processing in the brain and thus can capture disease mechanisms, is a promising emerging approach that is only beginning to be utilized in treatment response prediction. The predictors extracted via such models could improve interpretability, which is critical for clinical decision-making. We also identify several methodological limitations across the reviewed studies and provide suggestions for addressing them. Namely, we consider problems with dichotomizing treatment outcomes, the importance of investigating more than one treatment in a given study for differential treatment response predictions, the need for a patient-centered approach for defining treatment outcomes, and finally, the use of internal and external validation methods for improving model generalizability.

Predicting treatment response using EEG in major depressive disorder: A machine-learning meta-analysis

Selecting a course of treatment in psychiatry remains a trial-and-error process, and this long-standing clinical challenge has prompted an increased focus on predictive models of treatment response using machine learning techniques. Electroencephalography (EEG) represents a cost-effective and scalable potential measure to predict treatment response to major depressive disorder. We performed separate meta-analyses to determine the ability of models to distinguish between responders and non-responders using EEG across treatments, as well as a performed subgroup analysis of response to transcranial magnetic stimulation (rTMS), and antidepressants (Registration Number: CRD42021257477) in Major Depressive Disorder by searching PubMed, Scopus, and Web of Science for articles published between January 1960 and February 2022. We included 15 studies that predicted treatment responses among patients with major depressive disorder using machine-learning techniques. Within a random-effects model with a restricted maximum likelihood estimator comprising 758 patients, the pooled accuracy across studies was 83.93% (95% CI: 78.90-89.29), with an Area-Under-the-Curve (AUC) of 0.850 (95% CI: 0.747-0.890), and partial AUC of 0.779. The average sensitivity and specificity across models were 77.96% (95% CI: 60.05-88.70), and 84.60% (95% CI: 67.89-92.39), respectively. In a subgroup analysis, greater performance was observed in predicting response to rTMS (Pooled accuracy: 85.70% (95% CI: 77.45-94.83), Area-Under-the-Curve (AUC): 0.928, partial AUC: 0.844), relative to antidepressants (Pooled accuracy: 81.41% (95% CI: 77.45-94.83, AUC: 0.895, pAUC: 0.821). Furthermore, across all meta-analyses, the specificity (true negatives) of EEG models was greater than the sensitivity (true positives), suggesting that EEG models thus far better identify non-responders than responders to treatment in MDD. Studies varied widely in important features across models, although relevant features included absolute and relative power in frontal and temporal electrodes, measures of connectivity, and asymmetry across hemispheres. Predictive models of treatment response using EEG hold promise in major depressive disorder, although there is a need for prospective model validation in independent datasets, and a greater emphasis on replicating physiological markers. Crucially, standardization in cut-off values and clinical scales for defining clinical response and non-response will aid in the reproducibility of findings and the clinical utility of predictive models. Furthermore, several models thus far have used data from open-label trials with small sample sizes and evaluated performance in the absence of training and testing sets, which increases the risk of statistical overfitting. Large consortium studies are required to establish predictive signatures of treatment response using EEG, and better elucidate the replicability of specific markers. Additionally, it is speculated that greater performance was observed in rTMS models, since EEG is assessing neural networks more likely to be directly targeted by rTMS, comprising electrical activity primarily near the surface of the cortex. Prospectively, there is a need for models that examine the comparative effectiveness of multiple treatments across the same patients. However, this will require a thoughtful consideration towards cumulative treatment effects, and whether washout periods between treatments should be utilised. Regardless, longitudinal cross-over trials comparing multiple treatments across the same group of patients will be an important prerequisite step to both facilitate precision psychiatry and identify generalizable physiological predictors of response between and across treatment options.

A novel EEG-based major depressive disorder detection framework with two-stage feature selection

Background

Major depressive disorder (MDD) is a common mental illness, characterized by persistent depression, sadness, despair, etc., troubling people’s daily life and work seriously.

Methods

In this work, we present a novel automatic MDD detection framework based on EEG signals. First of all, we derive highly MDD-correlated features, calculating the ratio of extracted features from EEG signals at frequency bands between \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\beta$$\end{document}β and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha$$\end{document}α. Then, a two-stage feature selection method named PAR is presented with the sequential combination of Pearson correlation coefficient (PCC) and recursive feature elimination (RFE), where the advantages lie in minimizing the feature searching space. Finally, we employ widely used machine learning methods of support vector machine (SVM), logistic regression (LR), and linear regression (LNR) for MDD detection with the merit of feature interpretability.

Results

Experiment results show that our proposed MDD detection framework achieves competitive results. The accuracy and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$F_{1}$$\end{document}F1 score are up to 0.9895 and 0.9846, respectively. Meanwhile, the regression determination coefficient \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R^2$$\end{document}R2 for MDD severity assessment is up to 0.9479. Compared with existing MDD detection methods with the best accuracy of 0.9840 and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$F_1$$\end{document}F1 score of 0.97, our proposed framework achieves the state-of-the-art MDD detection performance.

Conclusions

Development of this MDD detection framework can be potentially deployed into a medical system to aid physicians to screen out MDD patients.

The two decades brainclinics research archive for insights in neurophysiology (TDBRAIN) database

In neuroscience, electroencephalography (EEG) data is often used to extract features (biomarkers) to identify neurological or psychiatric dysfunction or to predict treatment response. At the same time neuroscience is becoming more data-driven, made possible by computational advances. In support of biomarker development and methodologies such as training Artificial Intelligent (AI) networks we present the extensive Two Decades-Brainclinics Research Archive for Insights in Neurophysiology (TDBRAIN) EEG database. This clinical lifespan database (5–89 years) contains resting-state, raw EEG-data complemented with relevant clinical and demographic data of a heterogenous collection of 1274 psychiatric patients collected between 2001 to 2021. Main indications included are Major Depressive Disorder (MDD; N = 426), attention deficit hyperactivity disorder (ADHD; N = 271), Subjective Memory Complaints (SMC: N = 119) and obsessive-compulsive disorder (OCD; N = 75). Demographic-, personality- and day of measurement data are included in the database. Thirty percent of clinical and treatment outcome data will remain blinded for prospective validation and replication purposes. The TDBRAIN database and code are available on the Brainclinics Foundation website at www.brainclinics.com/resources and on Synapse at www.synapse.org/TDBRAIN.

Measurement(s)	psychiatric and healthy resting state EEG
Technology Type(s)	electro-encephalography (EEG)
Sample Characteristic - Organism	human
Sample Characteristic - Environment	laboratory environment
Sample Characteristic - Location	Province (The Netherlands)

A benchmark for prediction of psychiatric multimorbidity from resting EEG data in a large pediatric sample

Psychiatric disorders are among the most common and debilitating illnesses across the lifespan and begin usually during childhood and adolescence, which emphasizes the importance of studying the developing brain. Most of the previous pediatric neuroimaging studies employed traditional univariate statistics on relatively small samples. Multivariate machine learning approaches have a great potential to overcome the limitations of these approaches. On the other hand, the vast majority of existing multivariate machine learning studies have focused on differentiating between children with an isolated psychiatric disorder and typically developing children. However, this line of research does not reflect the real-life situation as the majority of children with a clinical diagnosis have multiple psychiatric disorders (multimorbidity), and consequently, a clinician has the task to choose between different diagnoses and/or the combination of multiple diagnoses. Thus, the goal of the present benchmark is to predict psychiatric multimorbidity in children and adolescents. For this purpose, we implemented two kinds of machine learning benchmark challenges: The first challenge targets the prediction of the seven most prevalent DSM-V psychiatric diagnoses for the available data set, of which each individual can exhibit multiple ones concurrently (i.e. multi-task multi-label classification). Based on behavioral and cognitive measures, a second challenge focuses on predicting psychiatric symptom severity on a dimensional level (i.e. multiple regression task). For the present benchmark challenges, we will leverage existing and future data from the biobank of the Healthy Brain Network (HBN) initiative, which offers a unique large-sample dataset (N = 2042) that provides a wide array of different psychiatric developmental disorders and true hidden data sets. Due to limited real-world practicability and economic viability of MRI measurements, the present challenge will permit only resting state EEG data and demographic information to derive predictive models. We believe that a community driven effort to derive predictive markers from these data using advanced machine learning algorithms can help to improve the diagnosis of psychiatric developmental disorders.

The Sample Size Matters: To What Extent the Participant Reduction Affects the Outcomes of a Neuroscientific Research. A Case-Study in Neuromarketing Field

The sample size is a crucial concern in scientific research and even more in behavioural neurosciences, where besides the best practice it is not always possible to reach large experimental samples. In this study we investigated how the outcomes of research change in response to sample size reduction. Three indices computed during a task involving the observations of four videos were considered in the analysis, two related to the brain electroencephalographic (EEG) activity and one to autonomic physiological measures, i.e., heart rate and skin conductance. The modifications of these indices were investigated considering five subgroups of sample size (32, 28, 24, 20, 16), each subgroup consisting of 630 different combinations made by bootstrapping n (n = sample size) out of 36 subjects, with respect to the total population (i.e., 36 subjects). The correlation analysis, the mean squared error (MSE), and the standard deviation (STD) of the indexes were studied at the participant reduction and three factors of influence were considered in the analysis: the type of index, the task, and its duration (time length). The findings showed a significant decrease of the correlation associated to the participant reduction as well as a significant increase of MSE and STD (p < 0.05). A threshold of subjects for which the outcomes remained significant and comparable was pointed out. The effects were to some extents sensitive to all the investigated variables, but the main effect was due to the task length. Therefore, the minimum threshold of subjects for which the outcomes were comparable increased at the reduction of the spot duration.

Advanced Signal Processing Methods for Characterization of Schizophrenia

OBJECTIVE

Schizophrenia is a severe mental disorder associated with nerobiological deficits. Auditory oddball P300 have been found to be one of the most consistent markers of schizophrenia. The goal of this study is to find quantitative features that can objectively distinguish patients with schizophrenia (SCZs) from healthy controls (HCs) based on their recorded auditory odd-ball P300 electroencephalogram (EEG) data.

METHODS

Using EEG dataset, we develop a machine learning (ML) algorithm to distinguish 57 SCZs from 66 HCs. The proposed ML algorithm has three steps. In the first step, a brain source localization (BSL) procedure using the linearly constrained minimum variance (LCMV) beamforming approach is employed on EEG signals to extract source waveforms from 30 specified brain regions. In the second step, a method for estimating effective connectivity, referred to as symbolic transfer entropy (STE), is applied to the source waveforms. In the third step the ML algorithm is applied to the STE connectivity matrix to determine whether a set of features can be found that successfully discriminate SCZ from HC.

RESULTS

The findings revealed that the SCZs have significantly higher effective connectivity compared to HCs and the selected STE features could achieve an accuracy of 92.68%, with a sensitivity of 92.98% and specificity of 92.42%.

CONCLUSION

The findings imply that the extracted features are from the regions that are mainly affected by SCZ and can be used to distinguish SCZs from HCs.

SIGNIFICANCE

The proposed ML algorithm may prove to be a promising tool for the clinical diagnosis of schizophrenia.

Inclusion of genetic variants in an ensemble of gradient boosting decision trees does not improve the prediction of citalopram treatment response

Identifying in advance who is unlikely to respond to a specific antidepressant treatment is crucial to precision medicine efforts. The current work leverages genome-wide genetic variation and machine learning to predict response to the antidepressant citalopram using data from the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial (n = 1257 with both valid genomic and outcome data). A confirmatory approach selected 11 SNPs previously reported to predict response to escitalopram in a sample different from the current study. A novel exploratory approach selected SNPs from across the genome using nested cross-validation with elastic net logistic regression with a predominantly lasso penalty (alpha = 0.99). SNPs from each approach were combined with baseline clinical predictors and treatment response outcomes were predicted using a stacked ensemble of gradient boosting decision trees. Using pre-treatment clinical and symptom predictors only, out-of-fold prediction of a novel treatment response definition based on STAR*D treatment guidelines was acceptable, AUC = .659, 95% CI [0.629, 0.689]. The inclusion of SNPs using confirmatory or exploratory selection methods did not improve the out-of-fold prediction of treatment response (AUCs were .662, 95% CI [0.632, 0.692] and .655, 95% CI [0.625, 0.685], respectively). A similar pattern of results were observed for the secondary outcomes of the presence or absence of distressing side effects regardless of treatment response and achieving remission or satisfactory partial response, assuming medication tolerance. In the current study, incorporating SNP variation into prognostic models did not enhance the prediction of citalopram response in the STAR*D sample.

Artificial Intelligence for Mental Health Care: Clinical Applications, Barriers, Facilitators, and Artificial Wisdom

Artificial intelligence (AI) is increasingly employed in health care fields such as oncology, radiology, and dermatology. However, the use of AI in mental health care and neurobiological research has been modest. Given the high morbidity and mortality in people with psychiatric disorders, coupled with a worsening shortage of mental health care providers, there is an urgent need for AI to help identify high-risk individuals and provide interventions to prevent and treat mental illnesses. While published research on AI in neuropsychiatry is rather limited, there is a growing number of successful examples of AI's use with electronic health records, brain imaging, sensor-based monitoring systems, and social media platforms to predict, classify, or subgroup mental illnesses as well as problems such as suicidality. This article is the product of a study group held at the American College of Neuropsychopharmacology conference in 2019. It provides an overview of AI approaches in mental health care, seeking to help with clinical diagnosis, prognosis, and treatment, as well as clinical and technological challenges, focusing on multiple illustrative publications. Although AI could help redefine mental illnesses more objectively, identify them at a prodromal stage, personalize treatments, and empower patients in their own care, it must address issues of bias, privacy, transparency, and other ethical concerns. These aspirations reflect human wisdom, which is more strongly associated than intelligence with individual and societal well-being. Thus, the future AI or artificial wisdom could provide technology that enables more compassionate and ethically sound care to diverse groups of people.

Major Depressive Disorder Classification Based on Different Convolutional Neural Network Models: Deep Learning Approach

The human brain is characterized by complex structural, functional connections that integrate unique cognitive characteristics. There is a fundamental hurdle for the evaluation of both structural and functional connections of the brain and the effects in the diagnosis and treatment of neurodegenerative diseases. Currently, there is no clinically specific diagnostic biomarker capable of confirming the diagnosis of major depressive disorder (MDD). Therefore, exploring translational biomarkers of mood disorders based on deep learning (DL) has valuable potential with its recently underlined promising outcomes. In this article, an electroencephalography (EEG)-based diagnosis model for MDD is built through advanced computational neuroscience methodology coupled with a deep convolutional neural network (CNN) approach. EEG recordings are analyzed by modeling 3 different deep CNN structure, namely, ResNet-50, MobileNet, Inception-v3, in order to dichotomize MDD patients and healthy controls. EEG data are collected for 4 main frequency bands (Δ, θ, α, and β, accompanying spatial resolution with location information by collecting data from 19 electrodes. Following the pre-processing step, different DL architectures were employed to underline discrimination performance by comparing classification accuracies. The classification performance of models based on location data, MobileNet architecture generated 89.33% and 92.66% classification accuracy. As to the frequency bands, delta frequency band outperformed compared to other bands with 90.22% predictive accuracy and area under curve (AUC) value of 0.9 for ResNet-50 architecture. The main contribution of the study is the delineation of distinctive spatial and temporal features using various DL architectures to dichotomize 46 MDD subjects from 46 healthy subjects. Exploring translational biomarkers of mood disorders based on DL perspective is the main focus of this study and, though it is challenging, with its promising potential to improve our understanding of the psychiatric disorders, computational methods are highly worthy for the diagnosis process and valuable in terms of both speed and accuracy compared with classical approaches.

Classification of Depression Through Resting-State Electroencephalogram as a Novel Practice in Psychiatry: Review

Background

Machine learning applications in health care have increased considerably in the recent past, and this review focuses on an important application in psychiatry related to the detection of depression. Since the advent of computational psychiatry, research based on functional magnetic resonance imaging has yielded remarkable results, but these tools tend to be too expensive for everyday clinical use.

Objective

This review focuses on an affordable data-driven approach based on electroencephalographic recordings. Web-based applications via public or private cloud-based platforms would be a logical next step. We aim to compare several different approaches to the detection of depression from electroencephalographic recordings using various features and machine learning models.

Methods

To detect depression, we reviewed published detection studies based on resting-state electroencephalogram with final machine learning, and to predict therapy outcomes, we reviewed a set of interventional studies using some form of stimulation in their methodology.

Results

We reviewed 14 detection studies and 12 interventional studies published between 2008 and 2019. As direct comparison was not possible due to the large diversity of theoretical approaches and methods used, we compared them based on the steps in analysis and accuracies yielded. In addition, we compared possible drawbacks in terms of sample size, feature extraction, feature selection, classification, internal and external validation, and possible unwarranted optimism and reproducibility. In addition, we suggested desirable practices to avoid misinterpretation of results and optimism.

Conclusions

This review shows the need for larger data sets and more systematic procedures to improve the use of the solution for clinical diagnostics. Therefore, regulation of the pipeline and standard requirements for methodology used should become mandatory to increase the reliability and accuracy of the complete methodology for it to be translated to modern psychiatry.

Use and performance of machine learning models for type 2 diabetes prediction in community settings: A systematic review and meta-analysis

OBJECTIVE

We aimed to identify machine learning (ML) models for type 2 diabetes (T2DM) prediction in community settings and determine their predictive performance.

METHOD

Systematic review of ML predictive modelling studies in 13 databases since 2009 was conducted. Primary outcomes included metrics of discrimination, calibration, and classification. Secondary outcomes included important variables, level of validation, and intended use of models. Meta-analysis of c-indices, subgroup analyses, meta-regression, publication bias assessments and sensitivity analyses were conducted.

RESULTS

Twenty-three studies (40 prediction models) were included. Studies with high-, moderate-, and low- risk of bias were 3, 14, and 6 respectively. All studies conducted internal validation whereas none conducted external validation of their models. Twenty studies provided classification metrics to varying extents whereas only 7 studies performed model calibration. Eighteen studies reported information on both the variables used for model development and the feature importance. Twelve studies highlighted potential applicability of their models for T2DM screening. Meta-analysis produced a good pooled c-index (0.812). Sources of heterogeneity were identified through subgroup analyses and meta-regression. Issues pertaining to methodological quality and reporting were observed.

CONCLUSIONS

We found evidence of good performance of ML models for T2DM prediction in the community. Improvements to methodology, reporting and validation are needed before they can be used at scale.

Evaluation of a Machine Learning Model Based on Pretreatment Symptoms and Electroencephalographic Features to Predict Outcomes of Antidepressant Treatment in Adults With Depression: A Prespecified Secondary Analysis of a Randomized Clinical Trial

IMPORTANCE

Despite the high prevalence and potential outcomes of major depressive disorder, whether and how patients will respond to antidepressant medications is not easily predicted.

OBJECTIVE

To identify the extent to which a machine learning approach, using gradient-boosted decision trees, can predict acute improvement for individual depressive symptoms with antidepressants based on pretreatment symptom scores and electroencephalographic (EEG) measures.

DESIGN, SETTING, AND PARTICIPANTS

This prognostic study analyzed data collected as part of the International Study to Predict Optimized Treatment in Depression, a randomized, prospective open-label trial to identify clinically useful predictors and moderators of response to commonly used first-line antidepressant medications. Data collection was conducted at 20 sites spanning 5 countries and including 518 adult outpatients (18-65 years of age) from primary care or specialty care practices who received a diagnosis of current major depressive disorder between December 1, 2008, and September 30, 2013. Patients were antidepressant medication naive or willing to undergo a 1-week washout period of any nonprotocol antidepressant medication. Statistical analysis was conducted from January 5 to June 30, 2019.

EXPOSURES

Participants with major depressive disorder were randomized in a 1:1:1 ratio to undergo 8 weeks of treatment with escitalopram oxalate (n = 162), sertraline hydrochloride (n = 176), or extended-release venlafaxine hydrochloride (n = 180).

MAIN OUTCOMES AND MEASURES

The primary objective was to predict improvement in individual symptoms, defined as the difference in score for each of the symptoms on the 21-item Hamilton Rating Scale for Depression from baseline to week 8, evaluated using the C index.

RESULTS

The resulting data set contained 518 patients (274 women; mean [SD] age, 39.0 [12.6] years; mean [SD] 21-item Hamilton Rating Scale for Depression score improvement, 13.0 [7.0]). With the use of 5-fold cross-validation for evaluation, the machine learning model achieved C index scores of 0.8 or higher on 12 of 21 clinician-rated symptoms, with the highest C index score of 0.963 (95% CI, 0.939-1.000) for loss of insight. The importance of any single EEG feature was higher than 5% for prediction of 7 symptoms, with the most important EEG features being the absolute delta band power at the occipital electrode sites (O1, 18.8%; Oz, 6.7%) for loss of insight. Over and above the use of baseline symptom scores alone, the use of both EEG and baseline symptom features was associated with a significant increase in the C index for improvement in 4 symptoms: loss of insight (C index increase, 0.012 [95% CI, 0.001-0.020]), energy loss (C index increase, 0.035 [95% CI, 0.011-0.059]), appetite changes (C index increase, 0.017 [95% CI, 0.003-0.030]), and psychomotor retardation (C index increase, 0.020 [95% CI, 0.008-0.032]).

CONCLUSIONS AND RELEVANCE

This study suggests that machine learning may be used to identify independent associations of symptoms and EEG features to predict antidepressant-associated improvements in specific symptoms of depression. The approach should next be prospectively validated in clinical trials and settings.

TRIAL REGISTRATION

ClinicalTrials.gov Identifier: NCT00693849.

Predicting treatment effects in unipolar depression: A meta-review

There is increasing interest in clinical prediction models in psychiatry, which focus on developing multivariate algorithms to guide personalized diagnostic or management decisions. The main target of these models is the prediction of treatment response to different antidepressant therapies. This is because the ability to predict response based on patients' personal data may allow clinicians to make improved treatment decisions, and to provide more efficacious or more tolerable medications to the right patient. We searched the literature for systematic reviews about treatment prediction in the context of existing treatment modalities for adult unipolar depression, until July 2019. Treatment effect is defined broadly to include efficacy, safety, tolerability and acceptability outcomes. We first focused on the identification of individual predictor variables that might predict treatment response, and second, we considered multivariate clinical prediction models. Our meta-review included a total of 10 systematic reviews; seven (from 2014 to 2018) focusing on individual predictor variables and three focusing on clinical prediction models. These identified a number of sociodemographic, phenomenological, clinical, neuroimaging, remote monitoring, genetic and serum marker variables as possible predictor variables for treatment response, alongside statistical and machine-learning approaches to clinical prediction model development. Effect sizes for individual predictor variables were generally small and clinical prediction models had generally not been validated in external populations. There is a need for rigorous model validation in large external data-sets to prove the clinical utility of models. We also discuss potential future avenues in the field of personalized psychiatry, particularly the combination of multiple sources of data and the emerging field of artificial intelligence and digital mental health to identify new individual predictor variables.

Use of Machine Learning for Predicting Escitalopram Treatment Outcome From Electroencephalography Recordings in Adult Patients With Depression

IMPORTANCE

Social and economic costs of depression are exacerbated by prolonged periods spent identifying treatments that would be effective for a particular patient. Thus, a tool that reliably predicts an individual patient's response to treatment could significantly reduce the burden of depression.

OBJECTIVE

To estimate how accurately an outcome of escitalopram treatment can be predicted from electroencephalographic (EEG) data on patients with depression.

DESIGN, SETTING, AND PARTICIPANTS

This prognostic study used a support vector machine classifier to predict treatment outcome using data from the first Canadian Biomarker Integration Network in Depression (CAN-BIND-1) study. The CAN-BIND-1 study comprised 180 patients (aged 18-60 years) diagnosed with major depressive disorder who had completed 8 weeks of treatment. Of this group, 122 patients had EEG data recorded before the treatment; 115 also had EEG data recorded after the first 2 weeks of treatment.

INTERVENTIONS

All participants completed 8 weeks of open-label escitalopram (10-20 mg) treatment.

MAIN OUTCOMES AND MEASURES

The ability of EEG data to predict treatment outcome, measured as accuracy, specificity, and sensitivity of the classifier at baseline and after the first 2 weeks of treatment. The treatment outcome was defined in terms of change in symptom severity, measured by the Montgomery-Åsberg Depression Rating Scale, before and after 8 weeks of treatment. A patient was designated as a responder if the Montgomery-Åsberg Depression Rating Scale score decreased by at least 50% during the 8 weeks and as a nonresponder if the score decrease was less than 50%.

RESULTS

Of the 122 participants who completed a baseline EEG recording (mean [SD] age, 36.3 [12.7] years; 76 [62.3%] female), the classifier was able to identify responders with an estimated accuracy of 79.2% (sensitivity, 67.3%; specificity, 91.0%) when using only the baseline EEG data. For a subset of 115 participants who had additional EEG data recorded after the first 2 weeks of treatment, use of these data increased the accuracy to 82.4% (sensitivity, 79.2%; specificity, 85.5%).

CONCLUSIONS AND RELEVANCE

These findings demonstrate the potential utility of EEG as a treatment planning tool for escitalopram therapy. Further development of the classification tools presented in this study holds the promise of expediting the search for optimal treatment for each patient.

Entropy analysis of heart rate variability and its application to recognize major depressive disorder: A pilot study

BACKGROUND

The current method to evaluate major depressive disorder (MDD) relies on subjective clinical interviews and self-questionnaires.

OBJECTIVE

Autonomic imbalance in MDD patients is characterized using entropy measures of heart rate variability (HRV). A machine learning approach for screening depression based on the entropy is demonstrated.

METHODS

The participants experience five experimental phases: baseline (BASE), stress task (MAT), stress task recovery (REC1), relaxation task (RLX), and relaxation task recovery (REC2). The four entropy indices, approximate entropy, sample entropy, fuzzy entropy, and Shannon entropy, are extracted for each phase, and a total of 20 features are used. A support vector machine classifier and recursive feature elimination are employed for classification.

RESULTS

The entropy features are lower in the MDD group; however, the disease does not have a significant effect. Experimental tasks significantly affect the features. The entropy did not recover during REC1. The differences in the entropy features between the two groups increased after MAT and showed the largest gap in REC2. We achieved 70% accuracy, 64% sensitivity, and 76% specificity with three optimal features during RLX and REC2.

CONCLUSION

Monitoring of HRV complexity changes when a subject experiences autonomic arousal and recovery can potentially facilitate objective depression recognition.

Internet-Based Management for Depressive Disorder

The advances in the Internet and related technologies may lead to changes in professional roles of psychiatrists and psychotherapists. The application of artificial intelligence (AI) and electronic measurement-based care (eMBC) in the treatment of depressive disorder has addressed more interest. AI could play a role in population health management and patient administration as well as assist physicians to make a decision in the real-world clinical practice. The eMBC strengthens MBC through web/mobile devices and telephone consulting services, to monitor disease progression, and customizes the MBC interface in electronic medical record systems (EMRs).

Can Machine Learning help us in dealing with treatment resistant depression? A review

BACKGROUND

About one third of patients treated with antidepressant do not show sufficient symptoms relief and up to 15% of patients remain symptomatic even after multiple trials are applied, configuring a state called treatment resistant depression (TRD). A clear definition of this state and the understanding of underlying mechanisms contributing to chronic disability caused by major depressive disorder is still unknown. Therefore, Machine Learning (ML) techniques emerged in the last years as interesting approaches to deal with such complex problems.

METHODS

We performed a bibliographic search on Pubmed, Google Scholar and Medline of clinical, imaging, genetic and EEG ML classification studies on treatment-responding depression and TRD as well as studies trying to predict response to a specific treatment in already established TRD. The inclusion criteria were met by eleven studies. Seven focused on the definition of predictors of TRD onset while four attempted to predict the response to specific treatments in TRD.

RESULTS

The results showed that it seems possible to classify between responders MDD and TRD with good accuracies based on clinical variables. Moreover, some studies reported the possibility of using EEG measures to predict response to different pharmacological and non-pharmacological treatments in established TRD.

LIMITATIONS

The definition of TRD, the selection of variables together with ML algorithms and pipelines varies across the studies, ultimately determining the unfeasibility to implement these models in clinical practice.

CONCLUSIONS

The findings suggest that ML could be a valid approach to increase our understanding of TRD and to better classify and stratify this disorder, which may ultimately help clinicians in the assessment of major depressive disorders.

An animal health example of managing and analyzing a large volume of data on a PC: Modeling body weight and age of over 13 million cats for explanatory and predictive purposes

Large amounts of animal health data are available to researchers, but are often stored in different formats and information silos. Analysis of this existing information can provide new insights into the health and welfare of animals and possibly reduce the need to collect additional data. The objective of this study was to develop a method of managing and analyzing large amounts of data on a personal computer that can be run within 24 h to limit the time and resources spent deploying models on larger servers. This paper describes an overall approach that makes use of existing methods for data acquisition and modeling, but adapts and combines them in a way that allows manipulation and analysis of large volumes of data on a PC. This included a total of five steps: removing errors; removing data points outside the scope of a specific hypothesis; creating descriptive statistics; developing explanatory and/or predictive models; and assessing the fit or accuracy of the models created. The approach was developed using electronic medical records for 19,416,753 feline patients from 3972 anonymized veterinary clinics in the United States and Canada, recorded between January 1981 and June 2016. Data regarding patient signalment (age, sex, breed, reproductive status) and body weight were extracted from the records and used to create linear regression models to describe body weight in cats of different ages, breeds, genders and reproductive status. Ordinary least squares linear regression and stochastic gradient descent linear regression were compared to determine their effectiveness and suitability for creating predictive models with large datasets, using 10 fold cross validation. This approach could be used to build workflows to create models to determine exploratory and predictive properties of health parameters for animals and people. The ability to work with large datasets on a PC or equivalent technology was demonstrated. Significant interactions were present among sex, reproductive status and age. A peak in weight occurred between 6 and 9 years depending on the sex, reproductive status and breed. The predictive ability of the two models was similar, with both producing a root mean square error of 1.45 and a mean absolute error of 1.09, and mean error that was approximately zero on the validation dataset.

A machine learning ensemble to predict treatment outcomes following an Internet intervention for depression

BACKGROUND

Some Internet interventions are regarded as effective treatments for adult depression, but less is known about who responds to this form of treatment.

METHOD

An elastic net and random forest were trained to predict depression symptoms and related disability after an 8-week course of an Internet intervention, Deprexis, involving adults (N = 283) from across the USA. Candidate predictors included psychopathology, demographics, treatment expectancies, treatment usage, and environmental context obtained from population databases. Model performance was evaluated using predictive R2$\lpar R_{{\rm pred}}^2\rpar\comma $ the expected variance explained in a new sample, estimated by 10 repetitions of 10-fold cross-validation.

RESULTS

An ensemble model was created by averaging the predictions of the elastic net and random forest. Model performance was compared with a benchmark linear autoregressive model that predicted each outcome using only its baseline. The ensemble predicted more variance in post-treatment depression (8.0% gain, 95% CI 0.8-15; total $R_{{\rm pred}}^2 \; $= 0.25), disability (5.0% gain, 95% CI -0.3 to 10; total $R_{{\rm pred}}^2 \; $= 0.25), and well-being (11.6% gain, 95% CI 4.9-19; total $R_{{\rm pred}}^2 \; $= 0.29) than the benchmark model. Important predictors included comorbid psychopathology, particularly total psychopathology and dysthymia, low symptom-related disability, treatment credibility, lower access to therapists, and time spent using certain Deprexis modules.

CONCLUSION

A number of variables predict symptom improvement following an Internet intervention, but each of these variables makes relatively small contributions. Machine learning ensembles may be a promising statistical approach for identifying the cumulative contribution of many weak predictors to psychosocial depression treatment response.

Prediction of rTMS treatment response in major depressive disorder using machine learning techniques and nonlinear features of EEG signal

BACKGROUND

Prediction of therapeutic outcome of repetitive transcranial magnetic stimulation (rTMS) treatment is an important purpose that eliminates financial and psychological consequences of applying inefficient therapy. To achieve this goal we proposed a method based on machine learning to classify responders (R) and non- responders (NR) to rTMS treatment for major depression disorder (MDD) patients.

METHODS

19 electrodes resting state EEG was recorded from 46 MDD patients before treatment. Then patients underwent 7 weeks of rTMS, and 23 of them responded to treatment. Features extracted from EEG include Lempel-Ziv complexity (LZC), Katz fractal dimension (KFD), correlation dimension (CD), the power spectral density, features based on bispectrum, frontal and prefrontal cordance and combination of them. The most relevant features were selected by the minimal-redundancy-maximal-relevance (mRMR) feature selection algorithm. For classifying two groups of R and NR, k-nearest neighbors (KNN) were applied. The performance of the proposed method was evaluated by leave-1-out cross-validation. For further study, the capability of features in differentiating R and NR was investigated by a statistical test.

RESULTS

Effective EEG features for prediction of rTMS treatment response were found. EEG beta power, the sum of bispectrum diagonal elements in delta and beta bands and CD were the most discriminative features. Power of beta classified R and NR with the high performance of 91.3% accuracy, 91.3% specificity, and 91.3% sensitivity.

LIMITATIONS

Lack of large sample size restricted our method for using in clinical applications.

CONCLUSION

This considerable high accuracy indicates that our proposed method with power and some of the nonlinear and bispectral features can lead to promising results in predicting treatment outcome of rTMS for MDD patients only by one session pretreatment EEG recording.

Electroencephalography in Psychiatric Surgery: Past Use and Future Directions

The last two decades have seen a re-emergence of surgery for intractable psychiatric disease, in large part due to increased use of deep brain stimulation. The development of more precise, image-guided, less invasive interventions has improved the safety of these procedures, even though the relative merits of modulation at various targets remain under investigation. With an increase in the number and type of interventions for modulating mood/anxiety circuits, the need for biomarkers to guide surgeries and predict treatment response is as critical as ever. Electroencephalography (EEG) has a long history in clinical neurology, cognitive neuroscience, and functional neurosurgery, but has limited prior usage in psychiatric surgery. MEDLINE, Embase, and Psyc-INFO searches on the use of EEG in guiding psychiatric surgery yielded 611 articles, which were screened for relevance and quality. We synthesized three important themes. First, considerable evidence supports EEG as a biomarker for response to various surgical and non-surgical therapies, but large-scale investigations are lacking. Second, intraoperative EEG is likely more valuable than surface EEG for guiding target selection, but comes at the cost of greater invasiveness. Finally, EEG may be a promising tool for objective functional feedback in developing "closed-loop" psychosurgeries, but more systematic investigations are required.

Development of a point of care system for automated coma prognosis: a prospective cohort study protocol

INTRODUCTION

Coma is a deep state of unconsciousness that can be caused by a variety of clinical conditions. Traditional tests for coma outcome prediction are based mainly on a set of clinical observations. Recently, certain event-related potentials (ERPs), which are transient electroencephalogram (EEG) responses to auditory, visual or tactile stimuli, have been introduced as useful predictors of a positive coma outcome (ie, emergence). However, such tests require the skills of clinical neurophysiologists, who are not commonly available in many clinical settings. Additionally, none of the current standard clinical approaches have sufficient predictive accuracies to provide definitive prognoses.

OBJECTIVE

The objective of this study is to develop improved machine learning procedures based on EEG/ERP for determining emergence from coma.

METHODS AND ANALYSIS

Data will be collected from 50 participants in coma. EEG/ERP data will be recorded for 24 consecutive hours at a maximum of five time points spanning 30 days from the date of recruitment to track participants' progression. The study employs paradigms designed to elicit brainstem potentials, middle-latency responses, N100, mismatch negativity, P300 and N400. In the case of patient emergence, data are recorded on that occasion to form an additional basis for comparison. A relevant data set will be developed from the testing of 20 healthy controls, each spanning a 15-hour recording period in order to formulate a baseline. Collected data will be used to develop an automated procedure for analysis and detection of various ERP components that are salient to prognosis. Salient features extracted from the ERP and resting-state EEG will be identified and combined to give an accurate indicator of prognosis.

ETHICS AND DISSEMINATION

This study is approved by the Hamilton Integrated Research Ethics Board (project number 4840). Results will be disseminated through peer-reviewed journal articles and presentations at scientific conferences.

TRIAL REGISTRATION NUMBER

NCT03826407.

Use of machine learning in predicting clinical response to transcranial magnetic stimulation in comorbid posttraumatic stress disorder and major depression: A resting state electroencephalography study

BACKGROUND

Repetitive transcranial magnetic stimulation (TMS) is clinically effective for major depressive disorder (MDD) and investigational for other conditions including posttraumatic stress disorder (PTSD). Understanding the mechanisms of TMS action and developing biomarkers predicting response remain important goals. We applied a combination of machine learning and electroencephalography (EEG), testing whether machine learning analysis of EEG coherence would (1) predict clinical outcomes in individuals with comorbid MDD and PTSD, and (2) determine whether an individual had received a TMS course.

METHODS

We collected resting-state 8-channel EEG before and after TMS (5 Hz to the left dorsolateral prefrontal cortex). We used Lasso regression and Support Vector Machine (SVM) to test the hypothesis that baseline EEG coherence predicted the outcome and to assess if EEG coherence changed after TMS.

RESULTS

In our sample, clinical response to TMS were predictable based on pretreatment EEG coherence (n = 29). After treatment, 13/29 had more than 50% reduction in MDD self-report score 12/29 had more than 50% reduction in PTSD self-report score. For MDD, area under roc curve was for MDD was 0.83 (95% confidence interval 0.69-0.94) and for PTSD was 0.71 (95% confidence interval 0.54-0.87). SVM classifier was able to accurately assign EEG recordings to pre- and post-TMS treatment. The accuracy for Alpha, Beta, Theta and Delta bands was 75.4 ± 1.5%, 77.4 ± 1.4%, 73.8 ± 1.5%, and 78.6 ± 1.4%, respectively, all significantly better than chance (50%, p < 0.001).

LIMITATION

Limitations of this work include lack of sham condition, modest sample size, and a sparse electrode array. Despite these methodological limitations, we found validated and clinically meaningful results.

CONCLUSIONS

Machine learning successfully predicted non-response to TMS with high specificity, and identified pre- and post-TMS status using EEG coherence. This approach may provide mechanistic insights and may also become a clinically useful screening tool for TMS candidates.