Feasibility of machine learning classification of depression and anxiety symptoms among college students using 3D gait and sit-to-walk biomechanics.

Unveiling latent neural mechanisms of depressive symptoms: Resting-state EEG biomarkers from symptom network and source localization analysis.

Major depressive disorder (MDD) remains clinically diagnosed based on subjective symptoms rather than objective neurobiological markers, which limits diagnostic accuracy and the ability to tailor treatment. We present an ensemble hybrid framework that integrates graph neural networks (GNN) with unsupervised clustering to classify and subtype MDD using resting-state functional connectivity (rs-fMRI) profiles. A GNN was trained to distinguish MDD from healthy controls using functional connectivity derived brain graphs, and the resulting subject level embeddings were clustered to uncover subtype structure. We evaluated the approach on two public multisite cohorts, REST-meta-MDD (China; N = 1,604; 17 sites) and SRPBS (Japan; N = 446; 4 sites), using leave-one-site-out cross-validation and cross-national transfer. The classifier achieved 0.73 leave-one-site-out accuracy on REST-meta-MDD and retained 0.78 sensitivity when transferred from the Chinese to the Japanese cohort, outperforming BrainIB and CI GNN under the same protocol. To mitigate site related confounds, we applied a standardized preprocessing pipeline and ComBat harmonization. Clustering consistently identified three MDD subtypes with distinct connectivity signatures involving the default mode network and cerebellum, the insula-cingulum temporal circuit, and frontostriatal circuitry. These findings provide a reproducible and biologically interpretable stratification of MDD. Prospective studies will be needed to link these subtypes to treatment response and other clinically meaningful outcomes.

A lightweight variable-channel deep self-attention model for depression classification based on raw EEG signals

Depression is a prevalent mental health disorder that often goes undetected due to stigma and delayed clinical intervention. Social media offers a promising avenue for early detection by leveraging users’ posts, which reflect their emotional and psychological states. Most existing methods rely on datasets labeled via self-reporting, where users explicitly mention their diagnosis. This study explores whether segmenting user timelines around such self-reports–into pre- and post-report periods–can improve the classification of depressive versus control users. We evaluate multiple textual representations and assess the predictive value of different timeline segments. Experiments are conducted on several English and Spanish datasets, including one manually validated dataset that was initially labeled via self-reporting. Results indicate that temporal segmentation provides valuable insights and can enhance model performance. The proposed approach is also compared against state-of-the-art methods to assess its generalizability across languages and data conditions.

In adolescents, identifying objective biomarkers for treatment response is crucial for the development of effective interventions. Voice-based biomarkers have recently shown potential to capture treatment-related changes in Major Depressive Disorder (MDD). While prior studies have been cross-sectional experiments with single speech sample, this study addresses a critical gap by evaluating intra-patient changes in speech over treatment period, providing insight into how these voice biomarkers evolve within individuals. We collected pre- and post-treatment voice samples from 48 adolescent MDD patients. We hypothesized that deep learning models could detect clinically meaningful changes in depressive states during treatment. Therefore, we compared machine learning and deep learning models for depressive classification. Additionally, we introduced the Dual Voice-based Depressive State Analysis (DVDSA) method to categorize intra-patient depressive state changes as recovery, worsening, or unchanged, highlighting the deep learning models' ability to detect these variations. Among the acoustic features, only the fundamental frequency exhibits significant changes between pre- and post-treatment states after Holm-Bonferroni correction. Machine learning models demonstrate limited performance in distinguishing treatment states, with the best F1-score reaching 65.83%. In contrast, deep learning model, particularly WavLM, achieves remarkably higher performance in binary classification, with an F1-score of 78.05%. The WavLM maintains robust performance, when applied to the DVDSA method, achieves an F1-score of 70.58%. These findings suggest that machine learning models and individual acoustic features may not sufficiently capture treatment-related changes in MDD patients. This study underscores the value of deep learning models using the DVDSA method, addressing the limitations of pre- and post-treatment classification and highlighting their potential to advance personalized treatment strategies for adolescent MDD.

The detection and classification of depression and other mental disorders have become crucial in the modern era, particularly with the growing reliance on social media for self-expression. Existing systems often face challenges like limited prediction accuracy, difficulty forecasting future mental illnesses, and handling both clinical and non-clinical data. This study proposes a novel analytical model that not only screens individuals' current mental health status from social media content but also predicts the likelihood of future mental health issues. The proposed methodology integrates classical machine learning (ML) models, ensemble learning approaches, and pretrained models for enhanced detection and forecasting accuracy. The outcome shows that pre-trained language models accomplished maximized F1-score and overall performance significantly better than conventional ML and ensemble models. The system outperforms existing methods with a significant accuracy improvement, achieving 90.9% overall accuracy, a 7.2% improvement over traditional ML classifiers, 5.8% over ensemble models, and 11.3% over language models.

Artificial intelligence in depression diagnostics: A systematic review of methodologies and clinical applications.

Depression represents a major global health challenge, yet traditional clinical diagnosis faces limitations, including high costs, limited coverage, and low patient willingness. Social media platforms provide new opportunities for early depression screening through user-generated content. However, existing methods often lack systematic integration of clinical knowledge and fail to leverage multi-modal information comprehensively. We propose a DSM-5-guided methodology that systematically maps clinical diagnostic criteria to computable social media features across three modalities: textual semantics (BERT-based deep semantic extraction), behavioral patterns (temporal activity analysis), and topic distributions (LDA-based cognitive bias identification). We design a hierarchical architecture integrating BERT, Bi-LSTM, hierarchical attention, and multi-task learning to capture both character-level and post-level importance while jointly optimizing depression classification, symptom recognition, and severity assessment. Experiments on the WU3D dataset (32,570 users, 2.19 million posts) demonstrate that our model achieves 91.8% F1-score, significantly outperforming baseline methods (BERT: 85.6%, TextCNN: 78.6%, and SVM: 72.1%) and large language models (GPT-4 few-shot: 86.9%). Ablation studies confirm that each component contributes meaningfully with synergistic effects. The model provides interpretable predictions through attention visualization and outputs fine-grained symptom assessments aligned with DSM-5 criteria. With low computational cost (~50 ms inference time), local deployability, and superior privacy protection, our approach offers significant practical value for large-scale mental health screening applications. This work demonstrates that domain-specialized methods with explicit clinical knowledge integration remain highly competitive in the era of general-purpose large language models.

Background/Objectives: Depressive disorder (DD) is a prevalent psychiatric condition often diagnosed through subjective self-reports, which can be time-consuming and lead to inaccurate assessments. To enhance diagnostic precision, integrating Electroencephalography (EEG) with machine learning (ML) has gained attention as an objective approach for DD diagnosis and severity assessment. Methods: We propose an interpretable EEG-based ML framework that integrates optimized functional connectivity features, including Coherence, Phase Lag Index (PLI), and Granger causality, to explore EEG-based functional connectivity patterns in individuals clinically diagnosed with depressive DD and to model symptom severity and cognitive vulnerability. The identified biomarkers provide a promising foundation for developing objective, clinically actionable decision-support tools in psychiatric care. Feature selection was performed using the Neighborhood Component Analysis (NCA) method, and biomarkers were identified through statistical tests. Results: The highest classification performance (97.66% ± 2.05%accuracy, 99.20% ± 1.10% sensitivity, 95.91% ± 4.66% specificity, 98.00% ± 1.02% f1-score, and 0.95 ± 0.48 MCC) was achieved using 21 NCA-selected features with a KNN (K = 9) classifier. The best severity assessment (r2 = 0.89 ± 0.10, MSE = 3.96 ± 17.05) and cognitive impairment prediction (r2 = 0.89 ± 0.06, MSE = 0.23 ± 0.45) were obtained using an ANN regressor with 20 and 17 NCA-selected features, respectively. Conclusions: Our approach outperforms previous EEG-based ML models in DD classification and severity prediction using fewer features. Notably, this is the first study to use EEG connectivity features to predict patients' severity and cognitive impairment in DD. Coherence and PLI values from frontal and temporal pathways across the alpha, beta, and gamma sub-bands may serve as critical biomarkers for DD diagnosis, severity assessment, and prediction of cognitive impairment.

Although significant progress has been made in automatic diagnosis systems for depression, most of the work focuses on combining features from multiple modalities to improve classification accuracy, which generates a lot of space-time overhead and feature synchronization problems. This research work proposes a unimodal depression detection framework based on facial expression and facial motion features. Firstly, we propose a robust feature extraction method based on the ratio of facial landmark and theoretically prove that this feature has up-down, left-right translation, depth translation, rotation, and flip invariance. The features extracted based on this method maintain the topological structure relationship of facial landmarks in space and maintain the temporal correlation of frames before and after facial landmarks. Then, we provide a novel idea to solve the classification task of large-unit depression videos. The final depression classification result is obtained by decomposing the depression classification task of large-unit videos into the scoring task of multiple short-sequence units and then through the defined score aggregation function. Our key innovations include: (1) theoretically proven invariant facial landmark ratio features, (2) novel video decomposition into short-sequence units with pseudo-labeling, and (3) efficient SRTSNet architecture. On DAIC-WOZ dataset, our framework achieves F1 = 0.85, outperforming all unimodal methods and matching state-of-the-art multimodal approaches while using only facial features.

BackgroundAdvances in automated language and speech analysis using machine learning have validated digital biomarkers for non‐invasive detection of subtle cognitive changes. While distinguishing Alzheimer's Disease (AD) from Normal Controls (NC) is straightforward, classifying Mild Cognitive Impairment (MCI) remains challenging, due to its potential progression to AD or its association with other factors such as affective disorders, requiring detailed expert evaluation. Expanding upon prior research, this study assesses LANGaware's biomarker capabilities on recent data for (a) binary classification of AD versus NC, (b) multi‐class classification into AD, NC, and MCI, and (c) binary classification of Depression (D) versus NC.MethodDiagnoses have been collected for the above cases, provided by medical experts following standardized protocols. Participants were engaged in simple elicitation tasks, such as describing a picture or narrating an event. Digital biomarkers that reflect linguistic, speech, and acoustic features were extracted from recorded audio and corresponding transcripts. These biomarkers were then used as input for classification tasks, employing a tailored neural network and an XGBoost model to perform binary and multi‐class (three‐class) classifications. The methodology was designed to be easily adaptable to multiple languages.ResultFor cognitive classification, 15827 elicitation tasks (5173 AD, 7660 MCI, 2994 NC) from English‐ and Greek‐speaking participants were analyzed using nested cross‐validation. The binary classifier (AD vs. Healthy) achieved an average F1‐macro score of 89.01%, while the three‐class one (AD vs. MCI vs. NC) attained a score of 73.0%. For depression classification, 4421 elicitation tasks (1481 D, 2940 NC) from English‐speaking participants were evaluated, achieving an average F1‐macro score of 70.38%.ConclusionThe results obtained confirm the strong discriminative capability of the proposed biomarkers for the early detection of cognitive decline. The findings support the applicability of automated assessment, facilitating early diagnosis and timely intervention. Additionally, the depression classification experiment further complements the cognitive analysis, given the established link between depression and MCI. This study is crucial for ensuring the quality and reliability of the LANGaware product as it is increasingly adopted by diagnostic centers and hospitals. We express our gratitude to all organizations that contributed valuable data to this work.

Major depressive disorder (MDD) is a prevalent and debilitating mental illness, affecting over 264 million people worldwide and imposing substantial social and economic burdens. Identifying MDD using resting-state functional MRI (rs-fMRI) is a promising direction for early diagnosis and intervention. However, inter-site heterogeneity-arising from differences in scanners and acquisition protocols-poses a major obstacle to building generalizable models across imaging sites. To address this challenge, we propose the Unsupervised Joint Alignment (UJA) framework for cross-site MDD classification. To the best of our knowledge, this is one of the first attempts to explore unsupervised rs-fMRI adaptation using an adversarial learning-based approach that jointly aligns both domain-level distributions and class-level structures. Specifically, UJA employs a multi-head self-attention module to extract informative representations from rs-fMRI data, followed by a unified alignment scheme that integrates adversarial domain-wise alignment with class-wise alignment based on dual classifiers and the sliced Wasserstein distance. Extensive experiments on the REST-meta-MDD dataset demonstrate that UJA consistently outperforms existing comparative methods across multiple cross-site scenarios. Ablation studies further highlight the complementary benefits of the dual alignment strategy. These findings highlight the potential of UJA to serve as a robust and generalizable tool for future clinical decision support in MDD diagnosis.

Depression is one of the most common mental disorders and one that has a great potential to affect people mentally, physically, and socially. Unfortunately, a majority of people either do not have access to treatment or avoid seeking help. In this context, many platforms have emerged to provide a space for discussion and support where users can interact anonymously. This study presents the results of our research on classifying depression in a specific community of religious people by analyzing texts posted on social media using semantic techniques, such as a comparative analysis of texts from users using ontologies. A supporting objective of the research was to create a natural language processing tool for classifying depression to obtain the corpus necessary for ontology creation. The resulting ontologies were analyzed and compared with each other and also with existing ontologies in the literature on general depression. The comparison was both qualitative and quantitative, taking into consideration similarity ratios for the quantitative comparison.

ObjectiveDepression represents a significant global health challenge, further complicated by the multifaceted and complex nature of its diagnosis and treatment. This study explores the application of multiple feature selection (FS) methodologies combined with XAI (explainable artificial intelligence) method named SHapley Additive exPlanations (SHAP) to enhance predictive accuracy in depression classification models using large-scale national survey data.MethodsLeveraging microdata from the National Mental Health Survey of Korea (2021), encompassing 5511 Korean adults, this research systematically evaluates how different FS-machine learning classifier combinations affect model performance and identifies nondiagnostic socioeconomic, psychological, and lifestyle factors associated with clinically diagnosed depression. By employing diverse FS methods (e.g., ReliefF, Markov Blanket, and Information Gain) across multiple machine learning classifiers, we systematically compare their performance across 12 classifiers.ResultsWe demonstrate that optimal FS method selection depends on machine learning classifier architecture, with ReliefF excelling in Stacking (F2-score =0.9851) and Markov Blanket performing best in ExtraTrees and LightGBM (F2-score =0.9848, 0.9838). After excluding core diagnostic criteria variables to avoid circularity, our analysis reveals that social distress (loneliness), reluctance to seek professional help, quality of life measures, and physical health comorbidities emerge as highly influential nondiagnostic predictors.ConclusionOur findings advance the field by: (1) systematically demonstrating that FS method effectiveness varies by machine learning classifier type, (2) providing a dual-layer XAI framework combining FS with SHAP for comprehensive interpretability, and (3) identifying culturally specific risk factors in an underrepresented Asian population using high-quality face-to-face collected data. These contributions provide methodological guidance for researchers developing interpretable depression prediction models and offer clinically actionable insights for identifying at-risk individuals in Korean populations.

MAMF-GCN model for anxious and non-anxious depression classification and neuroimaging marker recognition.

Acoustic voice analysis demonstrates potential as a non-invasive biomarker for depression, yet its generalizability across languages remains underexplored. This cross-sectional study aimed to identify a set of cross-culturally consistent acoustic features for depression screening using distinct Vietnamese and Japanese voice datasets. We analyzed anonymized recordings from 251 participants, comprising 123 Vietnamese individuals assessed via the self-report Beck Depression Inventory (BDI) and 128 Japanese individuals assessed via the clinician-rated Hamilton Depression Rating Scale (HAM-D). From 6373 features extracted with openSMILE, a multi-stage selection pipeline identified 12 cross-cultural features, primarily from the auditory spectrum (AudSpec), Mel-Frequency Cepstral Coefficients (MFCCs), and logarithmic Harmonics-to-Noise Ratio (logHNR) domains. The cross-cultural model achieved a combined Area Under the Curve (AUC) of 0.934, with performance disparities observed between the Japanese (AUC = 0.993) and Vietnamese (AUC = 0.913) cohorts. This disparity may be attributed to dataset heterogeneity, including mismatched diagnostic tools and differing sample compositions (clinical vs. mixed community). Furthermore, the limited number of high-risk cases (n = 33) warrants cautious interpretation regarding the reliability of reported AUC values for severe depression classification. These findings suggest the presence of a core acoustic signature related to physiological psychomotor changes that may transcend linguistic boundaries. This study advances the exploration of global vocal biomarkers but underscores the need for prospective, standardized multilingual trials to overcome the limitations of secondary data analysis.

In the modern world, people worldwide face different forms of depression due to factors such as workplace stress, economic pressures, and other causes. The rise of Artificial Intelligence (AI) has enabled data analysis and solving of real-world problems. People frequently use social media platforms to communicate and express their feelings. Hence, social media data is helpful for research purposes, particularly for automatic depression detection. Numerous scholarly works have explored using learning-based approaches to identify sadness from social media interactions. However, individual existing deep learning models have limitations, such as the inability to capture contextual and sequential dependencies in text fully. We addressed this by proposing a deep learning-based, non-invasive approach to identify depression in social media conversations. Our proposed approach involves a novel hybrid deep learning model, Depression Detection Network (DDNet), which combines Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) models. The model was trained and tested on a manually annotated dataset of 8500 depression-related tweets (6,800 for training and 1,700 for testing) collected via the Twitter Application Programming Interface (API). The DDNet model achieved a high accuracy of 96.21%, outperforming baseline models such as standalone LSTM (92.31%) and Recurrent Neural Network (RNN) (91.43%). Furthermore, we developed Hybrid Deep Learning-based Depression Detection (HDL-DD), an algorithm that processes social media text and predicts potential depressive tendencies. The experimental results indicate that DDNet significantly improves depression classification, achieving 95% precision, 96% recall, and 95% F1-score, demonstrating its effectiveness over existing methods. By recognizing depression with a 96.21% accuracy rate, our deep learning model outperformed previous state-of- the-art approaches, making it a promising tool for automated depression monitoring applications. This approach could be integrated into real-world social media-based mental health monitoring applications, supporting early intervention efforts and contributing to AI-driven healthcare solutions.

Utilizing Temporal Inductive Path Neural Networks for Accurate Voice-Based Depression Classification: A Detailed Approach for Analyzing Speech Patterns to Identify Mental Health States.

Depression is a frequent comorbidity among individuals with chronic diseases, amplifying morbidity and complicating disease management. In Bangladesh, data on the prevalence and predictors of depression in this population remain limited, particularly using advanced machine learning (ML) approaches. This cross-sectional study included 1,222 adult patients with clinically diagnosed chronic diseases, recruited from multiple healthcare centers between May and November 2024. Structured interviews collected information on sociodemographic, lifestyle, behavioral, clinical, and mental health-related factors. Probable depression was assessed with the Bangla version of the Patient Health Questionnaire-9 (PHQ-9). Traditional logistic regression and six ML algorithms were employed to identify factors associated with potential depression and evaluate model performance. SHAP and feature importance analyses were used to interpret ML results. The prevalence of probable depression among chronic disease patients was 29.7%. Adjusted regression analysis identified unemployment, urban residence, smokeless tobacco, alcohol and substance use, physical inactivity, short nighttime sleep duration (< 7 h), family history of chronic illness, and unmet mental healthcare needs as associated factors. CatBoost outperformed other ML models (accuracy: 71.1%; AUC: 0.76) in depression classification, with feature importance analyses consistently reporting residence, occupation, family history, and mental healthcare fulfillment as the top predictors. Depression is highly prevalent among patients with chronic diseases, shaped by a complex interplay of diverse factors. Machine learning models can accurately identify individuals at elevated risk and predictive factors, which can be used for targeted pre-screening and intervention strategies for this vulnerable population. Not applicable.