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

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.

Explainable multimodal prediction of treatment-resistance in patients with depression leveraging brain morphometry and natural language processing

Although 20 % of patients with depression receiving treatment do not achieve remission, predicting treatment-resistant depression (TRD) remains challenging. In this study, we aimed to develop an explainable multimodal prediction model for TRD using structured electronic medical record data, brain morphometry, and natural language processing. In total, 247 patients with a new depressive episode were included. TRD-predictive models were developed based on the combination of following parameters: selected tabular dataset features, independent components-map weightings from brain T1-weighted magnetic resonance imaging (MRI), and topic probabilities from clinical notes. All models applied the extreme gradient boosting (XGBoost) algorithm via five-fold cross-validation. The model using all data sources showed the highest area under the receiver operating characteristic of 0.794, followed by models that used combined brain MRI and structured data, brain MRI and clinical notes, clinical notes and structured data, brain MRI only, structured data only, and clinical notes only (0.770, 0.762, 0.728, 0.703, 0.684, and 0.569, respectively). Classifications of TRD were driven by several predictors, such as previous exposure to antidepressants and antihypertensive medications, sensorimotor network, default mode network, and somatic symptoms. Our findings suggest that a combination of clinical data with neuroimaging and natural language processing variables improves the prediction of TRD.

Reidentification of Participants in Shared Clinical Data Sets: Experimental Study

BACKGROUND

Large curated data sets are required to leverage speech-based tools in health care. These are costly to produce, resulting in increased interest in data sharing. As speech can potentially identify speakers (ie, voiceprints), sharing recordings raises privacy concerns. This is especially relevant when working with patient data protected under the Health Insurance Portability and Accountability Act.

OBJECTIVE

We aimed to determine the reidentification risk for speech recordings, without reference to demographics or metadata, in clinical data sets considering both the size of the search space (ie, the number of comparisons that must be considered when reidentifying) and the nature of the speech recording (ie, the type of speech task).

METHODS

Using a state-of-the-art speaker identification model, we modeled an adversarial attack scenario in which an adversary uses a large data set of identified speech (hereafter, the known set) to reidentify as many unknown speakers in a shared data set (hereafter, the unknown set) as possible. We first considered the effect of search space size by attempting reidentification with various sizes of known and unknown sets using VoxCeleb, a data set with recordings of natural, connected speech from >7000 healthy speakers. We then repeated these tests with different types of recordings in each set to examine whether the nature of a speech recording influences reidentification risk. For these tests, we used our clinical data set composed of recordings of elicited speech tasks from 941 speakers.

RESULTS

We found that the risk was inversely related to the number of comparisons an adversary must consider (ie, the search space), with a positive linear correlation between the number of false acceptances (FAs) and the number of comparisons (r=0.69; P<.001). The true acceptances (TAs) stayed relatively stable, and the ratio between FAs and TAs rose from 0.02 at 1 × 105 comparisons to 1.41 at 6 × 106 comparisons, with a near 1:1 ratio at the midpoint of 3 × 106 comparisons. In effect, risk was high for a small search space but dropped as the search space grew. We also found that the nature of a speech recording influenced reidentification risk, with nonconnected speech (eg, vowel prolongation: FA/TA=98.5; alternating motion rate: FA/TA=8) being harder to identify than connected speech (eg, sentence repetition: FA/TA=0.54) in cross-task conditions. The inverse was mostly true in within-task conditions, with the FA/TA ratio for vowel prolongation and alternating motion rate dropping to 0.39 and 1.17, respectively.

CONCLUSIONS

Our findings suggest that speaker identification models can be used to reidentify participants in specific circumstances, but in practice, the reidentification risk appears small. The variation in risk due to search space size and type of speech task provides actionable recommendations to further increase participant privacy and considerations for policy regarding public release of speech recordings.

Artificial Intelligence: A Game-Changer for Mental Health Care

Starting from the escalating global burden of mental health disorders, exacerbated by the COVID-19 pandemic, the article examines the potential of artificial intelligence (AI) to revolutionize mental health care. With nearly one in five adults facing mental health issues and suicide ranking as a leading cause of death among the young, the strained mental health system seeks innovative solutions. The text discusses the rapid evolution of AI, particularly in image analysis for early physical health diagnoses, and its promising applications in mental health, including predictive analytics for various disorders. AI's ability to analyze written language, speech characteristics, and physiological signals from wearables offers avenues for remote monitoring and early prognosis. Despite the need to address ethical considerations, particularly biases in data sets and concerns about potential patient detachment, the article advocates for AI as a complementary tool rather than a replacement for human involvement in mental health services. Overall, the article emphasizes the transformative potential of AI in enhancing diagnostics, monitoring, and treatment strategies for mental health disorders.

Predicting outcome with Intranasal Esketamine treatment: A machine-learning, three-month study in Treatment-Resistant Depression (ESK-LEARNING)

Treatment-resistant depression (TRD) represents a severe clinical condition with high social and economic costs. Esketamine Nasal Spray (ESK-NS) has recently been approved for TRD by EMA and FDA, but data about predictors of response are still lacking. Thus, a tool that can predict the individual patients' probability of response to ESK-NS is needed. This study investigates sociodemographic and clinical features predicting responses to ESK-NS in TRD patients using machine learning techniques. In a retrospective, multicentric, real-world study involving 149 TRD subjects, psychometric data (Montgomery-Asberg-Depression-Rating-Scale/MADRS, Brief-Psychiatric-Rating-Scale/BPRS, Hamilton-Anxiety-Rating-Scale/HAM-A, Hamilton-Depression-Rating-Scale/HAMD-17) were collected at baseline and at one month/T1 and three months/T2 post-treatment initiation. We trained three different random forest classifiers, able to predict responses to ESK-NS with accuracies of 68.53% at T1 and 66.26% at T2 and remission at T2 with 68.60% of accuracy. Features like severe anhedonia, anxious distress, mixed symptoms as well as bipolarity were found to positively predict response and remission. At the same time, benzodiazepine usage and depression severity were linked to delayed responses. Despite some limitations (i.e., retrospective study, lack of biomarkers, lack of a correct interrater-reliability across the different centers), these findings suggest the potential of machine learning in personalized intervention for TRD.

Predicting new-onset post-stroke depression from real-world data using machine learning algorithm

Introduction

Post-stroke depression (PSD) is a serious mental disorder after ischemic stroke. Early detection is important for clinical practice. This research aims to develop machine learning models to predict new-onset PSD using real-world data.

Methods

We collected data for ischemic stroke patients from multiple medical institutions in Taiwan between 2001 and 2019. We developed models from 61,460 patients and used 15,366 independent patients to test the models' performance by evaluating their specificities and sensitivities. The predicted targets were whether PSD occurred at 30, 90, 180, and 365 days post-stroke. We ranked the important clinical features in these models.

Results

In the study's database sample, 1.3% of patients were diagnosed with PSD. The average specificity and sensitivity of these four models were 0.83-0.91 and 0.30-0.48, respectively. Ten features were listed as important features related to PSD at different time points, namely old age, high height, low weight post-stroke, higher diastolic blood pressure after stroke, no pre-stroke hypertension but post-stroke hypertension (new-onset hypertension), post-stroke sleep-wake disorders, post-stroke anxiety disorders, post-stroke hemiplegia, and lower blood urea nitrogen during stroke.

Discussion

Machine learning models can provide as potential predictive tools for PSD and important factors are identified to alert clinicians for early detection of depression in high-risk stroke patients.

Early antidepressant treatment response prediction in major depression using clinical and TPH2 DNA methylation features based on machine learning approaches

OBJECTIVE

To identify DNA methylation and clinical features, and to construct machine learning classifiers to assign the patients with major depressive disorder (MDD) into responders and non-responders after a 2-week treatment into responders and non-responders.

METHOD

Han Chinese patients (291 in total) with MDD comprised the study population. Datasets contained demographic information, environment stress factors, and the methylation levels of 38 methylated sites of tryptophan hydroxylase 2 (TPH2) genes in peripheral blood samples. Recursive Feature Elimination (RFE) was employed to select features. Five classification algorithms (logistic regression, classification and regression trees, support vector machine, logitboost and random forests) were used to establish the models. Performance metrics (AUC, F-Measure, G-Mean, accuracy, sensitivity, specificity, positive predictive value and negative predictive value) were computed with 5-fold-cross-validation. Variable importance was evaluated by random forest algorithm.

RESULT

RF with RFE outperformed the other models in our samples based on the demographic information and clinical features (AUC = 61.2%, 95%CI: 60.1-62.4%) / TPH2 CpGs features (AUC = 66.6%, 95%CI: 65.4-67.8%) / both clinical and TPH2 CpGs features (AUC = 72.9%, 95%CI: 71.8-74.0%).

CONCLUSION

The effects of TPH2 on the early-stage antidepressant response were explored by machine learning algorithms. On the basis of the baseline depression severity and TPH2 CpG sites, machine learning approaches can enhance our ability to predict the early-stage antidepressant response. Some potentially important predictors (e.g., TPH2-10-60 (rs2129575), TPH2-2-163 (rs11178998), age of first onset, age) in early-stage treatment response could be utilized in future fundamental research, drug development and clinical practice.

Machine learning methods to predict outcomes of pharmacological treatment in psychosis

In recent years, machine learning (ML) has been a promising approach in the research of treatment outcome prediction in psychosis. In this study, we reviewed ML studies using different neuroimaging, neurophysiological, genetic, and clinical features to predict antipsychotic treatment outcomes in patients at different stages of schizophrenia. Literature available on PubMed until March 2022 was reviewed. Overall, 28 studies were included, among them 23 using a single-modality approach and 5 combining data from multiple modalities. The majority of included studies considered structural and functional neuroimaging biomarkers as predictive features used in ML models. Specifically, functional magnetic resonance imaging (fMRI) features contributed to antipsychotic treatment response prediction of psychosis with good accuracies. Additionally, several studies found that ML models based on clinical features might present adequate predictive ability. Importantly, by examining the additive effects of combining features, the predictive value might be improved by applying multimodal ML approaches. However, most of the included studies presented several limitations, such as small sample sizes and a lack of replication tests. Moreover, considerable clinical and analytical heterogeneity among included studies posed a challenge in synthesizing findings and generating robust overall conclusions. Despite the complexity and heterogeneity of methodology, prognostic features, clinical presentation, and treatment approaches, studies included in this review suggest that ML tools may have the potential to predict treatment outcomes of psychosis accurately. Future studies need to focus on refining feature characterization, validating prediction models, and evaluate their translation in real-world clinical practice.

EEG assessment of brain dysfunction for patients with chronic primary pain and depression under auditory oddball task

In 2019, the International Classification of Diseases 11th Revision International Classification of Diseases (ICD-11) put forward a new concept of "chronic primary pain" (CPP), a kind of chronic pain characterized by severe functional disability and emotional distress, which is a medical problem that deserves great attention. Although CPP is closely related to depressive disorder, its potential neural characteristics are still unclear. This paper collected EEG data from 67 subjects (23 healthy subjects, 22 patients with depression, and 22 patients with CPP) under the auditory oddball paradigm, systematically analyzed the brain network connection matrix and graph theory characteristic indicators, and classified the EEG and PLI matrices of three groups of people by frequency band based on deep learning. The results showed significant differences in brain network connectivity between CPP patients and depressive patients. Specifically, the connectivity within the frontoparietal network of the Theta band in CPP patients is significantly enhanced. The CNN classification model of EEG is better than that of PLI, with the highest accuracy of 85.01% in Gamma band in former and 79.64% in Theta band in later. We propose hyperexcitability in attentional control in CPP patients and provide a novel method for objective assessment of chronic primary pain.

Understanding treatment-resistant depression using "omics" techniques: A systematic review

BACKGROUND

Treatment-resistant depression (TRD) results in huge healthcare costs and poor patient clinical outcomes. Most studies have adopted a "candidate mechanism" approach to investigate TRD pathogenesis, however this is made more challenging due to the complex and heterogeneous nature of this condition. High-throughput "omics" technologies can provide a more holistic view and further insight into the underlying mechanisms involved in TRD development, expanding knowledge beyond already-identified mechanisms. This systematic review assessed the information from studies that examined TRD using hypothesis-free omics techniques.

METHODS

PubMed, MEDLINE, Embase, APA PsycInfo, Scopus and Web of Science databases were searched on July 2022. 37 human studies met the eligibility criteria, totalling 17,518 TRD patients, 571,402 healthy controls and 62,279 non-TRD depressed patients (including antidepressant responders and untreated MDD patients).

RESULTS

Significant findings were reported that implicate the role in TRD of various molecules, including polymorphisms, genes, mRNAs and microRNAs. The pathways most commonly reported by the identified studies were involved in immune system and inflammation, neuroplasticity, calcium signalling and neurotransmitters.

LIMITATIONS

Small sample sizes, variability in defining TRD, and heterogeneity in study design and methodology.

CONCLUSIONS

These findings provide insight into TRD pathophysiology, proposing future research directions for novel drug targets and potential biomarkers for clinical staging and response to antidepressants (citalopram/escitalopram in particular) and electroconvulsive therapy (ECT). Further validation is warranted in large prospective studies using standardised TRD criteria. A multi-omics and systems biology strategy with a collaborative effort will likely deliver robust findings for translation into the clinic.

Scoping Review on the Multimodal Classification of Depression and Experimental Study on Existing Multimodal Models

Depression is a prevalent comorbidity in patients with severe physical disorders, such as cancer, stroke, and coronary diseases. Although it can significantly impact the course of the primary disease, the signs of depression are often underestimated and overlooked. The aim of this paper was to review algorithms for the automatic, uniform, and multimodal classification of signs of depression from human conversations and to evaluate their accuracy. For the scoping review, the PRISMA guidelines for scoping reviews were followed. In the scoping review, the search yielded 1095 papers, out of which 20 papers (8.26%) included more than two modalities, and 3 of those papers provided codes. Within the scope of this review, supported vector machine (SVM), random forest (RF), and long short-term memory network (LSTM; with gated and non-gated recurrent units) models, as well as different combinations of features, were identified as the most widely researched techniques. We tested the models using the DAIC-WOZ dataset (original training dataset) and using the SymptomMedia dataset to further assess their reliability and dependency on the nature of the training datasets. The best performance was obtained by the LSTM with gated recurrent units (F1-score of 0.64 for the DAIC-WOZ dataset). However, with a drop to an F1-score of 0.56 for the SymptomMedia dataset, the method also appears to be the most data-dependent.

Repetitive Transcranial Magnetic Stimulation Improves Depression-like Behavior in Rats by Promoting Neural Stem Cell Proliferation and Differentiation

Repetitive transcranial magnetic stimulation (rTMS) is a novel non-invasive neuromodulation technique with neuroprotective properties and is used to treat depression. However, the underlying mechanism of action remains unclear. In this study, we examined the possible mechanism mediating the antidepressant effect of rTMS using animal experiments. Specific pathogen-free rats were treated with rTMS after exposure to social isolation combined with chronic unpredictable mild stress (CUMS). After four weeks of CUMS, the rats exhibited a significant decrease in spatial working memory assessed using open-field testing, a general loss of interest assessed with the sucrose preference test, and a significant reduction in spatial recognition memory ability assessed using the Y-maze. These behavioral deficits were accompanied by decreased numbers of astrocytes in the hippocampus, decreased expression of glial fibrillary acidic protein (GFAP), increased numbers of neural stem cells (NSCs), and increased expression of nestin protein. These results indicated that neuron damage occurred in the depression-like rats. After rTMS intervention, the depression-like behavior was alleviated significantly, and the numbers of NSCs and astrocytes, as well as the expression of GFAP and nestin proteins, returned to normal levels. Overall, it is likely that attenuation of NSC proliferation and differentiation into astrocytes produced a neuroprotective effect on hippocampal neurons, which might partly explain the mechanism by which rTMS alleviates depression.

Depression and Severity Detection Based on Body Kinematic Features: Using Kinect Recorded Skeleton Data of Simple Action

Relative limb movement is an important feature in assessing depression. In this study, we looked into whether a skeleton-mimetic task using natural stimuli may help people recognize depression. We innovatively used Kinect V2 to collect participant data. Sequential skeletal data was directly extracted from the original Kinect-3D and tetrad coordinates of the participant's 25 body joints. Two constructed skeletal datasets of whole-body joints (including binary classification and multi classification) were input into the proposed model for depression recognition after data preparation. We improved the temporal convolution network (TCN), creating novel spatial attention dilated TCN (SATCN) network that included a hierarchy of temporal convolution groups with different dilated convolution scales to capture important skeletal features and a spatial attention block for final result prediction. The depression and non-depression groups can be classified automatically with a maximum accuracy of 75.8% in the binary classification task, and 64.3% accuracy in the multi classification dataset to recognize more fine-grained identification of depression severity, according to experimental results. Our experiments and methods based on Kinect V2 can not only identify and screen depression patients but also effectively observe the recovery level of depression patients during the recovery process. For example, in the change from severe depression to moderate or mild depression multi classification dataset.

The potential of precision psychiatry: what is in reach?

Progress in developing personalised care for mental disorders is supported by numerous proof-of-concept machine learning studies in the area of risk assessment, diagnostics and precision prescribing. Most of these studies primarily use clinical data, but models might benefit from additional neuroimaging, blood and genetic data to improve accuracy. Combined, multimodal models might offer potential for stratification of patients for treatment. Clinical implementation of machine learning is impeded by a lack of wider generalisability, with efforts primarily focused on psychosis and dementia. Studies across all diagnostic groups should work to test the robustness of machine learning models, which is an essential first step to clinical implementation, and then move to prospective clinical validation. Models need to exceed clinicians' heuristics to be useful, and safe, in routine decision-making. Engagement of clinicians, researchers and patients in digitalisation and 'big data' approaches are vital to allow the generation and accessibility of large, longitudinal, prospective data needed for precision psychiatry to be applied into real-world psychiatric care.

Insula volumes in first-episode and chronic psychosis: A longitudinal MRI study

BACKGROUND

Alterations in insular grey matter (GM) volume has been consistently reported for affective and non-affective psychoses both in chronic and first-episode patients, ultimately suggesting that the insula might represent a good region to study in order to assess the longitudinal course of psychotic disorders. Therefore, in this longitudinal Magnetic Resonance Imaging (MRI) study, we aimed at further investigating the key role of insular volumes in psychosis.

MATERIAL AND METHODS

68 First-Episode Psychosis (FEP) patients, 68 patients with Schizophrenia (SCZ), 47 Bipolar Disorder (BD) patients, and 94 Healthy Controls (HC) were enrolled and underwent a 1.5 T MRI evaluation. A subsample of 99 subjects (10 HC, 23 BD, 29 SCZ, 37 FEP) was rescanned after 2,53 ± 1,68 years. The insular cortex was manually traced and then divided into an anterior and posterior portion. Group and correlation analyses were then performed both at baseline and at follow-up.

RESULTS

At baseline, greater anterior and lower posterior insular GM volumes were observed in chronic patients. At follow-up, we found that FEP patients had a significant GM volume increase from baseline to follow-up, especially in the posterior insula whereas chronic patients showed a relative stability. Finally, significant negative correlations between illness severity and pharmacological treatment and insular GM volumes were observed in the whole group of psychotic patients.

CONCLUSIONS

The longitudinal assessment of both chronic and first-episode patients allowed us to detect a complex pattern of GM abnormalities in selective sub-portions of insular volumes, ultimately suggesting that this structure could represent a key biological marker of psychotic disorders.

Improving Mental Health Services: A 50-Year Journey from Randomized Experiments to Artificial Intelligence and Precision Mental Health

This conceptual paper describes the current state of mental health services, identifies critical problems, and suggests how to solve them. I focus on the potential contributions of artificial intelligence and precision mental health to improving mental health services. Toward that end, I draw upon my own research, which has changed over the last half century, to highlight the need to transform the way we conduct mental health services research. I identify exemplars from the emerging literature on artificial intelligence and precision approaches to treatment in which there is an attempt to personalize or fit the treatment to the client in order to produce more effective interventions.

Psychiatry in the Digital Age: A Blessing or a Curse?

Social distancing and the shortage of healthcare professionals during the COVID-19 pandemic, the impact of population aging on the healthcare system, as well as the rapid pace of digital innovation are catalyzing the development and implementation of new technologies and digital services in psychiatry. Is this transformation a blessing or a curse for psychiatry? To answer this question, we conducted a literature review covering a broad range of new technologies and eHealth services, including telepsychiatry; computer-, internet-, and app-based cognitive behavioral therapy; virtual reality; digital applied games; a digital medicine system; omics; neuroimaging; machine learning; precision psychiatry; clinical decision support; electronic health records; physician charting; digital language translators; and online mental health resources for patients. We found that eHealth services provide effective, scalable, and cost-efficient options for the treatment of people with limited or no access to mental health care. This review highlights innovative technologies spearheading the way to more effective and safer treatments. We identified artificially intelligent tools that relieve physicians from routine tasks, allowing them to focus on collaborative doctor-patient relationships. The transformation of traditional clinics into digital ones is outlined, and the challenges associated with the successful deployment of digitalization in psychiatry are highlighted.

Optimizing prediction of response to antidepressant medications using machine learning and integrated genetic, clinical, and demographic data

Major depressive disorder (MDD) is complex and multifactorial, posing a major challenge of tailoring the optimal medication for each patient. Current practice for MDD treatment mainly relies on trial and error, with an estimated 42-53% response rates for antidepressant use. Here, we sought to generate an accurate predictor of response to a panel of antidepressants and optimize treatment selection using a data-driven approach analyzing combinations of genetic, clinical, and demographic factors. We analyzed the response patterns of patients to three antidepressant medications in the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study, and employed state-of-the-art machine learning (ML) tools to generate a predictive algorithm. To validate our results, we assessed the algorithm's capacity to predict individualized antidepressant responses on a separate set of 530 patients in STAR*D, consisting of 271 patients in a validation set and 259 patients in the final test set. This assessment yielded an average balanced accuracy rate of 72.3% (SD 8.1) and 70.1% (SD 6.8) across the different medications in the validation and test set, respectively (p < 0.01 for all models). To further validate our design scheme, we obtained data from the Pharmacogenomic Research Network Antidepressant Medication Pharmacogenomic Study (PGRN-AMPS) of patients treated with citalopram, and applied the algorithm's citalopram model. This external validation yielded highly similar results for STAR*D and PGRN-AMPS test sets, with a balanced accuracy of 60.5% and 61.3%, respectively (both p's < 0.01). These findings support the feasibility of using ML algorithms applied to large datasets with genetic, clinical, and demographic features to improve accuracy in antidepressant prescription.

Classification of first-episode psychosis using cortical thickness: A large multicenter MRI study

Machine learning classifications of first-episode psychosis (FEP) using neuroimaging have predominantly analyzed brain volumes. Some studies examined cortical thickness, but most of them have used parcellation approaches with data from single sites, which limits claims of generalizability. To address these limitations, we conducted a large-scale, multi-site analysis of cortical thickness comparing parcellations and vertex-wise approaches. By leveraging the multi-site nature of the study, we further investigated how different demographical and site-dependent variables affected predictions. Finally, we assessed relationships between predictions and clinical variables. 428 subjects (147 females, mean age 27.14) with FEP and 448 (230 females, mean age 27.06) healthy controls were enrolled in 8 centers by the ClassiFEP group. All subjects underwent a structural MRI and were clinically assessed. Cortical thickness parcellation (68 areas) and full cortical maps (20,484 vertices) were extracted. Linear Support Vector Machine was used for classification within a repeated nested cross-validation framework. Vertex-wise thickness maps outperformed parcellation-based methods with a balanced accuracy of 66.2% and an Area Under the Curve of 72%. By stratifying our sample for MRI scanner, we increased generalizability across sites. Temporal brain areas resulted as the most influential in the classification. The predictive decision scores significantly correlated with age at onset, duration of treatment, and positive symptoms. In conclusion, although far from the threshold of clinical relevance, temporal cortical thickness proved to classify between FEP subjects and healthy individuals. The assessment of site-dependent variables permitted an increase in the across-site generalizability, thus attempting to address an important machine learning limitation.

Predictors of change in suicidal ideation across treatment phases of major depressive disorder: analysis of the STAR*D data

The effects of common antidepressants on suicidal ideation (SI) is unclear. In the landmark STAR*D trial antidepressants were effective for Major Depressive Disorder (MDD) in early treatment phases, but less effective in later phases. The effects of antidepressants on SI across the entire sample of the STAR*D trial has never been investigated. We performed a secondary analysis of the STAR*D data with the primary outcome of change in score on the suicide item (item three) of the Hamilton Rating Scale for Depression (HRSD₁₇) across all four study levels. We used descriptive statistics and logistic regression analyses. Pearson correlation was used for change in SI versus change in depression (HRSD₁₆). Reduction in mean (SD) SI was greater in levels one: 0.29 (±0.78) (p < 0.001) and two: 0.26 (±0.88) (p < 0.001) than in levels three: 0.16 (±0.92) (p = 0.005) and four: 0.18 (±0.93) (p = 0.094). A history of past suicide attempts (OR 1.72, p = 0.007), comorbid medical illness (OR 2.23, p = 0.005), and a family history of drug abuse (OR 1.69, p = 0.008) were correlated with worsening of SI across level one. Treatment with bupropion (OR 0.24, p < 0.001) or buspirone (OR 0.24, p = 0.001) were correlated with lowering of SI across level two. Improvement in SI was correlated with improvement in overall depression (HRSD₁₆) at level one: r(3756) = 0.48; level two: r(1027) = 0.38; level three: r(249) = 0.31; and level four: r(75) = 0.42 (p < 0.001 for all levels). Improvement in SI is limited with pharmacotherapy in patients with treatment-resistant depression. Treatments with known anti-suicidal effects in MDD, such as ECT, should be considered in these patients.

Simple action for depression detection: using kinect-recorded human kinematic skeletal data

BACKGROUND

Depression, a common worldwide mental disorder, which brings huge challenges to family and social burden around the world is different from fluctuant emotion and psychological pressure in their daily life. Although body signs have been shown to present manifestations of depression in general, few researches focus on whole body kinematic cues with the help of machine learning methods to aid depression recognition. Using the Kinect V2 device to record participants' simple kinematic skeleton data of the participant's body joints, the presented spatial features and low-level features is directly extracted from the record original Kinect-3D coordinates. This research aimed to constructed machine learning model with the preprocessed data importing, which could be used for depression automatic classification.

METHODS

Considering some patients' conditions and current status and refer to psychiatrists' advices, simple and significant designed stimulus task will lead human skeleton data collection job. With original Kinect skeleton data extracting and preprocessing, the proposed experiment demonstrated four strong machine learning tools: Support Vector Machine, Logistic Regression, Random Forest and Gradient Boosting. Using the precision, recall, sensitivity, specificity, roc-curve, confusion matrix et.al, indicators were calculated as the measurement of methods, which were commonly used to evaluate classification methodologies.

RESULTS

Across screened 64 pairs with age and gender totally matching in depression and control group, and Gradient Boosting achieved the best performance with the prediction accuracy of 76.92%. Sorted by female (54.69%) and male for the gender-based depression recognition, we applied best performance classifier Gradient Boosting got prediction accuracy of 66.67% in the male group, and 71.73% in the female group. Utilizing the best model Gradient Boosting for age-based classification, prediction accuracy got 76.92% in the older group (age >40, 50% of total) and 53.85% accuracy in the younger group (age <= 40).

CONCLUSION

The depression and non-depression individuals can be well classified by computational models using Kinect captured skeletal data. The Gradient Boosting, an excellent machine learning tool, get the performance in the four methods we demonstrated. Meanwhile, in the gender-based depression classification also gets reasonable accuracy. In particular, the recognition results of the old group are significantly better than that of the young group. All these findings suggest that kinematic skeletal data based depression recognition can be applied as an effective tool for assisting in depression analysis.

Deep learning for the prediction of treatment response in depression

BACKGROUND

Mood disorders are characterized by heterogeneity in severity, symptoms and treatment response. The possibility of selecting the correct therapy on the basis of patient-specific biomarker may be a considerable step towards personalized psychiatry. Machine learning methods are gaining increasing popularity in the medical field. Once trained, the possibility to consider single patients in the analyses instead of whole groups makes them particularly appealing to investigate treatment response. Deep learning, a branch of machine learning, lately gained attention, due to its effectiveness in dealing with large neuroimaging data and to integrate them with clinical, molecular or -omics biomarkers.

METHODS

In this mini-review, we summarize studies that use deep learning methods to predict response to treatment in depression. We performed a bibliographic search on PUBMED, Google Scholar and Web of Science using the terms "psychiatry", "mood disorder", "depression", "treatment", "deep learning", "neural networks". Only studies considering patients' datasets are considered.

RESULTS

Eight studies met the inclusion criteria. Accuracies in prediction of response to therapy were considerably high in all studies, but results may be not easy to interpret.

LIMITATIONS

The major limitation for the current studies is the small sample size, which constitutes an issue for machine learning methods.

CONCLUSIONS

Deep learning shows promising results in terms of prediction of treatment response, often outperforming regression methods and reaching accuracies of around 80%. This could be of great help towards personalized medicine. However, more efforts are needed in terms of increasing datasets size and improved interpretability of results.

Sexual Regional Dimorphism of Post-Adolescent and Middle Age Brain Maturation. A Multi-center 3T MRI Study

Sex-related differences are tied into neurodevelopmental and lifespan processes, beginning early in the perinatal and developmental phases and continue into adulthood. The present study was designed to investigate sexual dimorphism of changes in gray matter (GM) volume in post-adolescence, with a focus on early and middle-adulthood using a structural magnetic resonance imaging (MRI) dataset of healthy controls from the European Network on Psychosis, Affective disorders and Cognitive Trajectory (ENPACT). Three hundred and seventy three subjects underwent a 3.0 T MRI session across four European Centers. Age by sex effects on GM volumes were investigated using voxel-based morphometry (VBM) and the Automated Anatomical Labeling atlas regions (ROI). Females and males showed overlapping and non-overlapping patterns of GM volume changes during aging. Overlapping age-related changes emerged in bilateral frontal and temporal cortices, insula and thalamus. Both VBM and ROI analyses revealed non-overlapping changes in multiple regions, including cerebellum and vermis, bilateral mid frontal, mid occipital cortices, left inferior temporal and precentral gyri. These findings highlight the importance of accounting for sex differences in cross-sectional analyses, not only in the study of normative changes, but particularly in the context of psychiatric and neurologic disorders, wherein sex effects may be confounded with disease-related changes.

Augmentation with Atypical Antipsychotics for Treatment-Resistant Depression

BACKGROUND

Treatment-resistant depression (TRD) is considered a common clinical condition often associated with relevant suicidal ideation and characterized by a severe functional impairment lifetime. Among the available drugs for the TRD treatment, second-generation antipsychotics (SGAs) have been reported as effective. In this context, the aim of this study was to review the clinical studies evaluating the efficacy of SGAs as add-on therapy in TRD.

METHODS

A comprehensive search on PubMed, Medline and PsychINFO of all randomized clinical trials (RCTs) assessing the augmentation with antipsychotics in TRD, published from January 2000 until March 2020, was performed. Sixteen RCTs studies met the inclusion criteria.

RESULTS

The reviewed studies showed that the add-on therapy with aripiprazole could be beneficial in the treatment of TRD. Furthermore, RCTs on quetiapine augmentation support its use in TRD, especially when comorbid anxiety or insomnia are present. The effects of risperidone and olanzapine as add-on in TRD were less studied, but preliminary data indicated an efficacy respect to placebo, making them a possible therapeutic option in TRD.

LIMITATIONS

The lack of consistency in the definition of TRD together with the small sample sizes and the heterogeneity of antipsychotics dosages used in the reviewed RCTs may have limited the strength of evidences obtained.

CONCLUSION

Overall, the available RCTs studies seem to support the hypothesis that the augmentation with SGAs, in particular aripiprazole and quetiapine, is a valid therapeutic option for TRD. However, to improve the therapeutic outcome of patients with TRD, larger and more homogeneous RCTs are needed.

Predicting second-generation antidepressant effectiveness in treating sadness using demographic and clinical information: A machine learning approach

INTRODUCTION

Current guidelines for choosing antidepressant medications involve a trial-and-error process. Most patients try multiple antidepressants before finding an effective antidepressant. This study uses demographic and clinical information to create models predicting effectiveness of different antidepressants in treating sadness in a nationally representative sample of US adults.

METHODS

A secondary analysis of the Collaborative Psychiatric Epidemiology Survey (CPES) was performed. Participants with or without a mental health diagnosis who reported sadness as a symptom, and were taking fluoxetine (n=156), sertraline (n=224), citalopram (n=91), paroxetine (n=156), venlafaxine (n=69), bupropion (n=92), or trazadone (n=26) within the past year were included. Two sets of principal component analyses (PCAs) and logistic regressions were performed: one determined associations between symptom clusters and antidepressant effectiveness for sadness, and the other created models to predict effectiveness. Both PCAs controlled for psychiatric and medical diagnoses, substance use, psychiatric medications, alternative treatments, and demographics.

RESULTS

Anxiety was associated with ineffectiveness of fluoxetine in treating sadness. Low mood scores were associated with ineffectiveness of paroxetine and venlafaxine, and fatigue was associated with ineffectiveness of sertraline. The models for predicting drug effectiveness had a mean accuracy of 83% and internal validity of 72%.

LIMITATIONS

CPES data were collected from 2001-2003, so newer drugs were not included. Effectiveness was for sadness, so results are not directly comparable to studies using overall depressive symptom reductions as outcomes.

CONCLUSION

Since fewer than 50% of patients currently respond to their first antidepressant, this model could provide modest improvement to choosing starting antidepressants in treating sadness.

Medical Instructed Real-Time Assistant for Patient with Glaucoma and Diabetic Conditions

Virtual assistants are involved in the daily activities of humans such as managing calendars, making appointments, and providing wake-up calls. They provide a conversational service to customers around-the-clock and make their daily life manageable. With this emerging trend, many well-known companies launched their own virtual assistants that manage the daily routine activities of customers. In the healthcare sector, virtual medical assistants also provide a list of relevant diseases linked to a specific symptom. Due to low accuracy and uncertainty, these generated recommendations are untrusted and may lead to hypochondriasis. In this study, we proposed a Medical Instructed Real-time Assistant (MIRA) that listens to the user’s chief complaint and predicts a specific disease. Instead of informing about the medical condition, the user is referred to a nearby appropriate medical specialist. We designed an architecture for MIRA that considers the limitations of existing virtual medical assistants such as weak authentication, lack of understanding multiple intent statements about a specific medical condition, and uncertain diagnosis recommendations. To implement the designed architecture, we collected the chief complaints along with the dialogue corpora of real patients. Then, we manually validated these data under the supervision of medical specialists. We then used these data for natural language understanding, disease identification, and appropriate response generation. For the prototype version of MIRA, we considered the cases of glaucoma (eye disease) and diabetes (an autoimmune disease) only. The performance measure of MIRA was evaluated in terms of accuracy (89%), precision (90%), sensitivity (89.8%), specificity (94.9%), and F-measure (89.8%). The task completion was calculated using Cohen’s Kappa ( k = 0.848 ) that categorizes MIRA as ‘Almost Perfect’. Furthermore, the voice-based authentication identifies the user effectively and prevent against masquerading attack. Simultaneously, the user experience shows relatively good results in all aspects based on the User Experience Questionnaire (UEQ) benchmark data. The experimental results show that MIRA efficiently predicts a disease based on chief complaints and supports the user in decision making.

Predictors of Treatment Resistance Across Different Clinical Subtypes of Depression: Comparison of Unipolar vs. Bipolar Cases

OBJECTIVE

Treatment-resistant depression (TRD) and treatment-resistant bipolar depression (TRBD) poses a significant clinical and societal burden, relying on different operational definitions and treatment approaches. The detection of clinical predictors of resistance is elusive, soliciting clinical subtyping of the depressive episodes, which represents the goal of the present study.

METHODS

A hundred and thirty-one depressed outpatients underwent psychopathological evaluation using major rating tools, including the Hamilton Rating Scale for Depression, which served for subsequent principal component analysis, followed-up by cluster analysis, with the ultimate goal to fetch different clinical subtypes of depression.

RESULTS

The cluster analysis identified two clinically interpretable, yet distinctive, groups among 53 bipolar (resistant cases = 15, or 28.3%) and 78 unipolar (resistant cases = 20, or 25.6%) patients. Among the MDD patients, cluster "1" included the following components: "Psychic symptoms, depressed mood, suicide, guilty, insomnia" and "genitourinary, gastrointestinal, weight loss, insight". Altogether, with broadly defined "mixed features," this latter cluster correctly predicted treatment outcome in 80.8% cases of MDD. The same "broadly-defined" mixed features of depression (namely, the standard Diagnostic and Statistical Manual for Mental Disorders, Fifth Edition-DSM-5-specifier plus increased energy, psychomotor activity, irritability) correctly classified 71.7% of BD cases, either as TRBD or not.

LIMITATIONS

Small sample size and high rate of comorbidity.

CONCLUSIONS

Although relying on different operational criteria and treatment history, TRD and TRBD seem to be consistently predicted by broadly defined mixed features among different clinical subtypes of depression, either unipolar or bipolar cases. If replicated by upcoming studies to encompass also biological and neuropsychological measures, the present study may aid in precision medicine and informed pharmacotherapy.