Correction: Evaluating the Clinical Feasibility of an Artificial Intelligence-Powered, Web-Based Clinical Decision Support System for the Treatment of Depression in Adults: Longitudinal Feasibility Study.

[This corrects the article DOI: 10.2196/31862.].

Towards Outcome-Driven Patient Subgroups: A Machine Learning Analysis Across Six Depression Treatment Studies.

BACKGROUND: Major depressive disorder (MDD) is a heterogeneous condition; multiple underlying neurobiological and behavioral substrates are associated with treatment response variability. Understanding the sources of this variability and predicting outcomes has been elusive. Machine learning (ML) shows promise in predicting treatment response in MDD, but its application is limited by challenges to the clinical interpretability of ML models, and clinicians often lack confidence in model results. In order to improve the interpretability of ML models in clinical practice, our goal was to demonstrate the derivation of treatment-relevant patient profiles comprised of clinical and demographic information using a novel ML approach. METHODS: We analyzed data from six clinical trials of pharmacological treatment for depression (total n = 5438) using the Differential Prototypes Neural Network (DPNN), a ML model that derives patient prototypes which can be used to derive treatment-relevant patient clusters while learning to generate probabilities for differential treatment response. A model classifying remission and outputting individual remission probabilities for five first-line monotherapies and three combination treatments was trained using clinical and demographic data. Prototypes were evaluated for interpretability by assessing differences in feature distributions (e.g. age, sex, symptom severity) and treatment-specific outcomes. RESULTS: A 3-prototype model achieved an area under the receiver operating curve of 0.66 and an expected absolute improvement in remission rate for those receiving the best predicted treatment of 6.5% (relative improvement of 15.6%) compared to the population remission rate. We identified three treatment-relevant patient clusters. Cluster A patients tended to be younger, to have increased levels of fatigue, and more severe symptoms. Cluster B patients tended to be older, female, have less severe symptoms, and the highest remission rates. Cluster C patients had more severe symptoms, lower remission rates, more psychomotor agitation, more intense suicidal ideation, and more somatic genital symptoms. CONCLUSION: It is possible to produce novel treatment-relevant patient profiles using ML models; doing so may improve interpretability of ML models and the quality of precision medicine treatments for MDD.

Cell type specific transcriptomic differences in depression show similar patterns between males and females but implicate distinct cell types and genes.

Major depressive disorder (MDD) is a common, heterogenous, and potentially serious psychiatric illness. Diverse brain cell types have been implicated in MDD etiology. Significant sexual differences exist in MDD clinical presentation and outcome, and recent evidence suggests different molecular bases for male and female MDD. We evaluated over 160,000 nuclei from 71 female and male donors, leveraging new and pre-existing single-nucleus RNA-sequencing data from the dorsolateral prefrontal cortex. Cell type specific transcriptome-wide threshold-free MDD-associated gene expression patterns were similar between the sexes, but significant differentially expressed genes (DEGs) diverged. Among 7 broad cell types and 41 clusters evaluated, microglia and parvalbumin interneurons contributed the most DEGs in females, while deep layer excitatory neurons, astrocytes, and oligodendrocyte precursors were the major contributors in males. Further, the Mic1 cluster with 38% of female DEGs and the ExN10_L46 cluster with 53% of male DEGs, stood out in the meta-analysis of both sexes.

Quantitative and carbon isotope ratio analysis of fatty acids isolated from human brain hemispheres.

Our knowledge surrounding the overall fatty acid profile of the adult human brain has been largely limited to extrapolations from brain regions in which the distribution of fatty acids varies. This is especially problematic when modeling brain fatty acid metabolism, therefore, an updated estimate of whole-brain fatty acid concentration is necessitated. Here, we sought to conduct a comprehensive quantitative analysis of fatty acids from entire well-characterized human brain hemispheres (n = 6) provided by the Douglas-Bell Canada Brain Bank. Additionally, exploratory natural abundance carbon isotope ratio (CIR; δ13 C, 13 C/12 C) analysis was performed to assess the origin of brain fatty acids. Brain fatty acid methyl esters (FAMEs) were quantified by gas chromatography (GC)-flame ionization detection and minor n-6 and n-3 polyunsaturated fatty acid pentafluorobenzyl esters by GC-mass spectrometry. Carbon isotope ratio values of identifiable FAMEs were measured by GC-combustion-isotope ratio mass spectrometry. Overall, the most abundant fatty acid in the human brain was oleic acid, followed by stearic acid (STA), palmitic acid (PAM), docosahexaenoic acid (DHA), and arachidonic acid (ARA). Interestingly, cholesterol as well as saturates including PAM and STA were most enriched in 13 C, while PUFAs including DHA and ARA were most depleted in 13 C. These findings suggest a contribution of endogenous synthesis utilizing dietary sugar substrates rich in 13 C, and a combination of marine, animal, and terrestrial PUFA sources more depleted in 13 C, respectively. These results provide novel insights on cerebral fatty acid origin and concentration, the latter serving as a valuable resource for future modeling of fatty acid metabolism in the human brain.

A systematic meta-review of patient-level predictors of psychological therapy outcome in major depressive disorder.

BACKGROUND: Psychological therapies are effective for treating major depressive disorder, but current clinical guidelines do not provide guidance on the personalization of treatment choice. Established predictors of psychotherapy treatment response could help inform machine learning models aimed at predicting individual patient responses to different therapy options. Here we sought to comprehensively identify known predictors. METHODS: EMBASE, Medline, PubMed, PsycINFO were searched for systematic reviews with or without meta-analysis published until June 2020 to identify individual patient-level predictors of response to psychological treatments. 3113 abstracts were identified and 300 articles assessed. We qualitatively synthesized our findings by predictor category (sociodemographic; symptom profile; social support; personality features; affective, cognitive, and behavioural; comorbidities; neuroimaging; genetics) and treatment type. We used the AMSTAR 2 to evaluate the quality of included reviews. RESULTS: Following screening and full-text assessment, 27 systematic reviews including 12 meta-analyses were eligible for inclusion. 74 predictors emerged for various psychological treatments, primarily cognitive behavioural therapy, interpersonal therapy, and mindfulness-based cognitive therapy. LIMITATIONS: A paucity of studies examining predictors of psychological treatment outcome, as well as methodological heterogeneities and publication biases limit the strength of the identified predictors. CONCLUSIONS: The synthesized predictors could be used to supplement clinical decision-making in selecting psychological therapies based on individual patient characteristics. These predictors could also be used as a priori input features for machine learning models aimed at predicting a given patient's likelihood of response to different treatment options for depression, and may contribute toward the development of patient-specific treatment recommendations in clinical guidelines.

Targeting microglia-oligodendrocyte crosstalk in neurodegenerative and psychiatric disorders.

To understand how various brain cell types communicate with each other to orchestrate functional processes, it is crucial to comprehend the signals used to relay such information. Therefore, an important challenge to studying complex brain diseases is to interrogate relevant interactions between cell types. The microglia-oligodendroglia interaction is an important example that has fundamental roles in physiological state and brain pathologies. Here, we review the latest findings on microglia-oligodendroglia interplay in physiological and pathological conditions. Furthermore, we provide an in silico ligand-receptor interaction analysis to explore potential druggable targets in multiple sclerosis (MS) and major depressive disorder (MDD).

Evaluating the perceived utility of an artificial intelligence-powered clinical decision support system for depression treatment using a simulation center.

Aifred is a clinical decision support system (CDSS) that uses artificial intelligence to assist physicians in selecting treatments for major depressive disorder (MDD) by providing probabilities of remission for different treatment options based on patient characteristics. We evaluated the utility of the CDSS as perceived by physicians participating in simulated clinical interactions. Twenty physicians who were either staff or residents in psychiatry or family medicine completed a study in which they had three 10-minute clinical interactions with standardized patients portraying mild, moderate, and severe episodes of MDD. During these scenarios, physicians were given access to the CDSS, which they could use in their treatment decisions. The perceived utility of the CDSS was assessed through self-report questionnaires, scenario observations, and interviews. 60% of physicians perceived the CDSS to be a useful tool in their treatment-selection process, with family physicians perceiving the greatest utility. Moreover, 50% of physicians would use the tool for all patients with depression, with an additional 35% noting that they would reserve the tool for more severe or treatment-resistant patients. Furthermore, clinicians found the tool to be useful in discussing treatment options with patients. The efficacy of this CDSS and its potential to improve treatment outcomes must be further evaluated in clinical trials.

Child abuse associates with increased recruitment of perineuronal nets in the ventromedial prefrontal cortex: a possible implication of oligodendrocyte progenitor cells.

Child abuse (CA) is a strong predictor of psychopathologies and suicide, altering normal trajectories of brain development in areas closely linked to emotional responses such as the prefrontal cortex (PFC). Yet, the cellular underpinnings of these enduring effects are unclear. Childhood and adolescence are marked by the protracted formation of perineuronal nets (PNNs), which orchestrate the closure of developmental windows of cortical plasticity by regulating the functional integration of parvalbumin interneurons into neuronal circuits. Using well-characterized post-mortem brain samples, we show that a history of CA is specifically associated with increased densities and morphological complexity of WFL-labeled PNNs in the ventromedial PFC (BA11/12), possibly suggesting increased recruitment and maturation of PNNs. Through single-nucleus sequencing and fluorescent in situ hybridization, we found that the expression of canonical components of PNNs is enriched in oligodendrocyte progenitor cells (OPCs), and that they are upregulated in CA victims. These correlational findings suggest that early-life adversity may lead to persistent patterns of maladaptive behaviors by reducing the neuroplasticity of cortical circuits through the enhancement of developmental OPC-mediated PNN formation.

Evaluating the Clinical Feasibility of an Artificial Intelligence-Powered, Web-Based Clinical Decision Support System for the Treatment of Depression in Adults: Longitudinal Feasibility Study.

BACKGROUND: Approximately two-thirds of patients with major depressive disorder do not achieve remission during their first treatment. There has been increasing interest in the use of digital, artificial intelligence-powered clinical decision support systems (CDSSs) to assist physicians in their treatment selection and management, improving the personalization and use of best practices such as measurement-based care. Previous literature shows that for digital mental health tools to be successful, the tool must be easy for patients and physicians to use and feasible within existing clinical workflows. OBJECTIVE: This study aims to examine the feasibility of an artificial intelligence-powered CDSS, which combines the operationalized 2016 Canadian Network for Mood and Anxiety Treatments guidelines with a neural network-based individualized treatment remission prediction. METHODS: Owing to the COVID-19 pandemic, the study was adapted to be completed entirely remotely. A total of 7 physicians recruited outpatients diagnosed with major depressive disorder according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition criteria. Patients completed a minimum of one visit without the CDSS (baseline) and 2 subsequent visits where the CDSS was used by the physician (visits 1 and 2). The primary outcome of interest was change in appointment length after the introduction of the CDSS as a proxy for feasibility. Feasibility and acceptability data were collected through self-report questionnaires and semistructured interviews. RESULTS: Data were collected between January and November 2020. A total of 17 patients were enrolled in the study; of the 17 patients, 14 (82%) completed the study. There was no significant difference in appointment length between visits (introduction of the tool did not increase appointment length; F2,24=0.805; mean squared error 58.08; P=.46). In total, 92% (12/13) of patients and 71% (5/7) of physicians felt that the tool was easy to use; 62% (8/13) of patients and 71% (5/7) of physicians rated that they trusted the CDSS. Of the 13 patients, 6 (46%) felt that the patient-clinician relationship significantly or somewhat improved, whereas 7 (54%) felt that it did not change. CONCLUSIONS: Our findings confirm that the integration of the tool does not significantly increase appointment length and suggest that the CDSS is easy to use and may have positive effects on the patient-physician relationship for some patients. The CDSS is feasible and ready for effectiveness studies. TRIAL REGISTRATION: ClinicalTrials.gov NCT04061642; http://clinicaltrials.gov/ct2/show/NCT04061642.

Fatty acid dysregulation in the anterior cingulate cortex of depressed suicides with a history of child abuse.

Child abuse (CA) strongly increases the lifetime risk of suffering from major depression and predicts an unfavorable course for the illness. Severe CA has been associated with a specific dysregulation of oligodendrocyte function and thinner myelin sheaths in the human anterior cingulate cortex (ACC) white matter. Given that myelin is extremely lipid-rich, it is plausible that these findings may be accompanied by a disruption of the lipid profile that composes the myelin sheath. This is important to explore since the composition of fatty acids (FA) in myelin phospholipids can influence its stability, permeability, and compactness. Therefore, the objective of this study was to quantify and compare FA concentrations in postmortem ACC white matter in the choline glycerophospholipid pool (ChoGpl), a key myelin phospholipid pool, between adult depressed suicides with a history of CA (DS-CA) matched depressed suicides without CA (DS) and healthy non-psychiatric controls (CTRL). Total lipids were extracted from 101 subjects according to the Folch method and separated into respective classes using thin-layer chromatography. FA methyl esters from the ChoGpl fraction were quantified using gas chromatography. Our analysis revealed specific effects of CA in FAs from the arachidonic acid synthesis pathway, which was further validated with RNA-sequencing data. Furthermore, the concentration of most FAs was found to decrease with age. By extending the previous molecular level findings linking CA with altered myelination in the ACC, these results provide further insights regarding white matter alterations associated with early-life adversity.

Identification of Suicidal Ideation in the Canadian Community Health Survey-Mental Health Component Using Deep Learning.

Introduction: Suicidal ideation (SI) is prevalent in the general population, and is a risk factor for suicide. Predicting which patients are likely to have SI remains challenging. Deep Learning (DL) may be a useful tool in this context, as it can be used to find patterns in complex, heterogeneous, and incomplete datasets. An automated screening system for SI could help prompt clinicians to be more attentive to patients at risk for suicide. Methods: Using the Canadian Community Health Survey-Mental Health Component, we trained a DL model based on 23,859 survey responses to classify patients with and without SI. Models were created to classify both lifetime SI and SI over the last 12 months. From 582 possible parameters we produced 96- and 21-feature versions of the models. Models were trained using an undersampling procedure that balanced the training set between SI and non-SI; validation was done on held-out data. Results: For lifetime SI, the 96 feature model had an Area under the receiver operating curve (AUC) of 0.79 and the 21 feature model had an AUC of 0.77. For SI in the last 12 months the 96 feature model had an AUC of 0.71 and the 21 feature model had an AUC of 0.68. In addition, sensitivity analyses demonstrated feature relationships in line with existing literature. Discussion: Although further study is required to ensure clinical relevance and sample generalizability, this study is an initial proof of concept for the use of DL to improve identification of SI. Sensitivity analyses can help improve the interpretability of DL models. This kind of model would help start conversations with patients which could lead to improved care and a reduction in suicidal behavior.

Age-related injury responses of human oligodendrocytes to metabolic insults: link to BCL-2 and autophagy pathways.

Myelin destruction and oligodendrocyte (OL) death consequent to metabolic stress is a feature of CNS disorders across the age spectrum. Using cells derived from surgically resected tissue, we demonstrate that young (

Using a simulation centre to evaluate preliminary acceptability and impact of an artificial intelligence-powered clinical decision support system for depression treatment on the physician-patient interaction.

BACKGROUND: Recently, artificial intelligence-powered devices have been put forward as potentially powerful tools for the improvement of mental healthcare. An important question is how these devices impact the physician-patient interaction. AIMS: Aifred is an artificial intelligence-powered clinical decision support system (CDSS) for the treatment of major depression. Here, we explore the use of a simulation centre environment in evaluating the usability of Aifred, particularly its impact on the physician-patient interaction. METHOD: Twenty psychiatry and family medicine attending staff and residents were recruited to complete a 2.5-h study at a clinical interaction simulation centre with standardised patients. Each physician had the option of using the CDSS to inform their treatment choice in three 10-min clinical scenarios with standardised patients portraying mild, moderate and severe episodes of major depression. Feasibility and acceptability data were collected through self-report questionnaires, scenario observations, interviews and standardised patient feedback. RESULTS: All 20 participants completed the study. Initial results indicate that the tool was acceptable to clinicians and feasible for use during clinical encounters. Clinicians indicated a willingness to use the tool in real clinical practice, a significant degree of trust in the system's predictions to assist with treatment selection, and reported that the tool helped increase patient understanding of and trust in treatment. The simulation environment allowed for the evaluation of the tool's impact on the physician-patient interaction. CONCLUSIONS: The simulation centre allowed for direct observations of clinician use and impact of the tool on the clinician-patient interaction before clinical studies. It may therefore offer a useful and important environment in the early testing of new technological tools. The present results will inform further tool development and clinician training materials.

Transcriptomic and clonal characterization of T cells in the human central nervous system.

T cells provide critical immune surveillance to the central nervous system (CNS), and the cerebrospinal fluid (CSF) is thought to be a main route for their entry. Further characterization of the state of T cells in the CSF in healthy individuals is important for understanding how T cells provide protective immune surveillance without damaging the delicate environment of the CNS and providing tissue-specific context for understanding immune dysfunction in neuroinflammatory disease. Here, we have profiled T cells in the CSF of healthy human donors and have identified signatures related to cytotoxic capacity and tissue adaptation that are further exemplified in clonally expanded CSF T cells. By comparing profiles of clonally expanded T cells obtained from the CSF of patients with multiple sclerosis (MS) and healthy donors, we report that clonally expanded T cells from the CSF of patients with MS have heightened expression of genes related to T cell activation and cytotoxicity.

Single-nucleus transcriptomics of the prefrontal cortex in major depressive disorder implicates oligodendrocyte precursor cells and excitatory neurons.

Major depressive disorder (MDD) has an enormous impact on global disease burden, affecting millions of people worldwide and ranking as a leading cause of disability for almost three decades. Past molecular studies of MDD employed bulk homogenates of postmortem brain tissue, which obscures gene expression changes within individual cell types. Here we used single-nucleus transcriptomics to examine ~80,000 nuclei from the dorsolateral prefrontal cortex of male individuals with MDD (n = 17) and of healthy controls (n = 17). We identified 26 cellular clusters, and over 60% of these showed differential gene expression between groups. We found that the greatest dysregulation occurred in deep layer excitatory neurons and immature oligodendrocyte precursor cells (OPCs), and these contributed almost half (47%) of all changes in gene expression. These results highlight the importance of dissecting cell-type-specific contributions to the disease and offer opportunities to identify new avenues of research and novel targets for treatment.

Developmental trajectory of oligodendrocyte progenitor cells in the human brain revealed by single cell RNA sequencing.

Characterizing the developmental trajectory of oligodendrocyte progenitor cells (OPC) is of great interest given the importance of these cells in the remyelination process. However, studies of human OPC development remain limited by the availability of whole cell samples and material that encompasses a wide age range, including time of peak myelination. In this study, we apply single cell RNA sequencing to viable whole cells across the age span and link transcriptomic signatures of oligodendrocyte-lineage cells with stage-specific functional properties. Cells were isolated from surgical tissue samples of second-trimester fetal, 2-year-old pediatric, 13-year-old adolescent, and adult donors by mechanical and enzymatic digestion, followed by percoll gradient centrifugation. Gene expression was analyzed using droplet-based RNA sequencing (10X Chromium). Louvain clustering analysis identified three distinct cellular subpopulations based on 5,613 genes, comprised of an early OPC (e-OPC) group, a late OPC group (l-OPC), and a mature OL (MOL) group. Gene ontology terms enriched for e-OPCs included cell cycle and development, for l-OPCs included extracellular matrix and cell adhesion, and for MOLs included myelination and cytoskeleton. The e-OPCs were mostly confined to the premyelinating fetal group, and the l-OPCs were most highly represented in the pediatric age group, corresponding to the peak age of myelination. Cells expressing a signature characteristic of l-OPCs were identified in the adult brain in situ using RNAScope. These findings highlight the transcriptomic variability in OL-lineage cells before, during, and after peak myelination and contribute to identifying novel pathways required to achieve remyelination.

A systematic meta-review of predictors of antidepressant treatment outcome in major depressive disorder.

INTRODUCTION: The heterogeneity of symptoms and complex etiology of depression pose a significant challenge to the personalization of treatment. Meanwhile, the current application of generic treatment approaches to patients with vastly differing biological and clinical profiles is far from optimal. Here, we conduct a meta-review to identify predictors of response to antidepressant therapy in order to select robust input features for machine learning models of treatment response. These machine learning models will allow us to learn associations between patient features and treatment response which have predictive value at the individual patient level; this learning can be optimized by selecting high-quality input features for the model. While current research is difficult to directly apply to the clinic, machine learning models built using knowledge gleaned from current research may become useful clinical tools. METHODS: The EMBASE and MEDLINE/PubMed online databases were searched from January 1996 to August 2017, using a combination of MeSH terms and keywords to identify relevant literature reviews. We identified a total of 1909 articles, wherein 199 articles met our inclusion criteria. RESULTS: An array of genetic, immune, endocrine, neuroimaging, sociodemographic, and symptom-based predictors of treatment response were extracted, varying widely in clinical utility. LIMITATIONS: Due to heterogeneous sample sizes, effect sizes, publication biases, and methodological disparities across reviews, we could not accurately assess the strength and directionality of every predictor. CONCLUSION: Notwithstanding our cautious interpretation of the results, we have identified a multitude of predictors that can be used to formulate a priori hypotheses regarding the input features for a computational model. We highlight the importance of large-scale research initiatives and clinically accessible biomarkers, as well as the need for replication studies of current findings. In addition, we provide recommendations for future improvement and standardization of research efforts in this field.

Analysis of Features Selected by a Deep Learning Model for Differential Treatment Selection in Depression.

Background: Deep learning has utility in predicting differential antidepressant treatment response among patients with major depressive disorder, yet there remains a paucity of research describing how to interpret deep learning models in a clinically or etiologically meaningful way. In this paper, we describe methods for analyzing deep learning models of clinical and demographic psychiatric data, using our recent work on a deep learning model of STAR*D and CO-MED remission prediction. Methods: Our deep learning analysis with STAR*D and CO-MED yielded four models that predicted response to the four treatments used across the two datasets. Here, we use classical statistics and simple data representations to improve interpretability of the features output by our deep learning model and provide finer grained understanding of their clinical and etiological significance. Specifically, we use representations derived from our model to yield features predicting both treatment non-response and differential treatment response to four standard antidepressants, and use linear regression and t-tests to address questions about the contribution of trauma, education, and somatic symptoms to our models. Results: Traditional statistics were able to probe the input features of our deep learning models, reproducing results from previous research, while providing novel insights into depression causes and treatments. We found that specific features were predictive of treatment response, and were able to break these down by treatment and non-response categories; that specific trauma indices were differentially predictive of baseline depression severity; that somatic symptoms were significantly different between males and females, and that education and low income proved important psycho-social stressors associated with depression. Conclusion: Traditional statistics can augment interpretation of deep learning models. Such interpretation can lend us new hypotheses about depression and contribute to building causal models of etiology and prognosis. We discuss dataset-specific effects and ideal clinical samples for machine learning analysis aimed at improving tools to assist in optimizing treatment.