Prospects for the Use of Machine Learning for Mood Disorders

Digital Diagnostics, Год журнала: 2025, Номер unknown

Опубликована: Янв. 28, 2025

Relevance. Mental disorders are one of the key medical and social issues. Over last years artificial intelligence (AI) methods including machine deep learning have been actively developing. This narrative review aimed to identify current promising areas for development application AI into clinical practice using example patients with depression bipolar disorder. Methods. The search publications was performed in January ─ February 2024 PubMed, Google Scholar, elibrary databases combination keywords: psychiatry, mental health, psychiatric disorder, depression, depressive episode, major learning, intelligence. included original articles on use devoted problems applying psychiatry published Russian or English 10 years. Results. Most often, neuroimaging (mainly MRI EEG), text, audio video data, electronic device molecular genetics, data its combination, used (ML) models mood disorders. Despite potential benefits implementation is currently challenging due number difficulties, such as small sample sizes, low representativeness, lack standardization, inclusion “noise” correlated variables models, model testing independent samples. Conclusion. Studies ML shown results early diagnosis affective episodes predicting response therapy. However, has a limitations, primarily insufficient validation. There need well-designed prospective cohort studies, well extensive high-quality capable identifying new relationships between order overcome these limitations.

Язык: Английский

---

Using effective connectivity-based predictive modeling to predict MDD scale scores from multisite rs-fMRI data

Journal of Neuroscience Methods, Год журнала: 2025, Номер 417, С. 110406 - 110406

Опубликована: Фев. 18, 2025

Язык: Английский

---

Advancing the diagnosis of major depressive disorder: Integrating neuroimaging and machine learning

World Journal of Psychiatry, Год журнала: 2025, Номер 15(3)

Опубликована: Фев. 26, 2025

Major depressive disorder (MDD), a psychiatric characterized by functional brain deficits, poses considerable diagnostic and treatment challenges, especially in adolescents owing to varying clinical presentations. Biomarkers hold substantial potential the field of mental health, enabling objective assessments physiological pathological states, facilitating early diagnosis, enhancing decision-making patient outcomes. Recent breakthroughs combine neuroimaging with machine learning (ML) distinguish activity patterns between MDD patients healthy controls, paving way for support personalized treatment. However, accuracy results depends on selection features algorithms. Ensuring privacy protection, ML model accuracy, fostering trust are essential steps prior implementation. Future research should prioritize establishment comprehensive legal frameworks regulatory mechanisms using diagnosis while safeguarding rights. By doing so, we can advance care MDD.

Язык: Английский

---

The application of artificial intelligence in insomnia, anxiety, and depression: A bibliometric analysis

Digital Health, Год журнала: 2025, Номер 11

Опубликована: Янв. 1, 2025

Mental health issues like insomnia, anxiety, and depression have increased significantly. Artificial intelligence (AI) has shown promise in diagnosing providing personalized treatment. This study aims to systematically review the application of AI addressing depression, identifying key research hotspots, forecasting future trends through bibliometric analysis. We analyzed a total 875 articles from Web Science Core Collection (2000-2024) using tools such as VOSviewer CiteSpace. These were used map trends, highlight international collaboration, examine contributions leading countries, institutions, authors field. The United States China lead field terms output collaborations. Key areas include "neural networks," "machine learning," "deep "human-robot interaction," particularly relation treatment approaches. However, challenges around data privacy, ethical concerns, interpretability models need be addressed. highlights growing role mental identifies priorities, improving quality, challenges, integrating more seamlessly into clinical practice. advancements will crucial global crisis.

Язык: Английский

---

A Framework for Comparison and Interpretation of Machine Learning Classifiers to Predict Autism on the ABIDE Dataset

Human Brain Mapping, Год журнала: 2025, Номер 46(5)

Опубликована: Март 17, 2025

ABSTRACT Autism is a neurodevelopmental condition affecting ~1% of the population. Recently, machine learning models have been trained to classify participants with autism using their neuroimaging features, though performance these varies in literature. Differences experimental setup hamper direct comparison different machine‐learning approaches. In this paper, five most widely used and best‐performing field were typically developing (TD) participants, functional connectivity matrices, structural volumetric measures, phenotypic information from Brain Imaging Data Exchange (ABIDE) dataset. Their was compared under same evaluation standard. The implemented included: graph convolutional networks (GCN), edge‐variational (EV‐GCN), fully connected (FCN), autoencoder followed by network (AE‐FCN) support vector (SVM). Our results show that all performed similarly, achieving classification accuracy around 70%. suggest inclusion criteria, data modalities, pipelines rather than may explain variations published highest our framework obtained when ensemble ( p < 0.001), leading an 72.2% AUC = 0.77 GCN classifiers. However, SVM classifier 70.1% 0.77, just marginally below GCN, significant differences not found comparing algorithms testing conditions > 0.05). Furthermore, we also investigated stability features identified SmoothGrad interpretation method. FCN model demonstrated selecting relevant contributing decision making. code available at https://github.com/YilanDong19/Machine‐learning‐with‐ABIDE .

Язык: Английский

---

Generalizability of clinical prediction models in mental health

Molecular Psychiatry, Год журнала: 2025, Номер unknown

Опубликована: Март 19, 2025

Abstract Concerns about the generalizability of machine learning models in mental health arise, partly due to sampling effects and data disparities between research cohorts real-world populations. We aimed investigate whether a model trained solely on easily accessible low-cost clinical can predict depressive symptom severity unseen, independent datasets from various contexts. This observational multi-cohort study included 3021 participants (62.03% females, M Age = 36.27 years, range 15–81) ten European settings, all diagnosed with an affective disorder. firstly compared inpatients same treatment center using 76 sociodemographic variables. An elastic net algorithm ten-fold cross-validation was then applied develop sparse for predicting depression based top five features (global functioning, extraversion, neuroticism, emotional abuse childhood, somatization). Model tested across nine external samples. The reliably predicted samples ( r 0.60, SD 0.089, p < 0.0001) each individual sample, ranging performance 0.48 general population sample 0.73 inpatients. These results suggest that have potential illness diverse offering insights could inform development more generalizable tools use routine psychiatric analysis.

Язык: Английский

---

Examining the Long-Term Impacts of Psychotropic Drugs and Considerations for People Discontinuing Treatment

Issues in Mental Health Nursing, Год журнала: 2025, Номер unknown, С. 1 - 9

Опубликована: Март 21, 2025

Psychotropic drugs dominate the mental healthcare landscape. This is despite contention over their proposed mechanism of action, concerns for adverse effects, and questionable effectiveness, especially long term. Mental health nurses are routinely involved in administering psychotropic drugs, observing managing providing information support to people prescribed these agents. critique explores current understanding action evidence effect burden implications term use. The role deprescribing supporting discontinue treatment considered.

Язык: Английский

---

Reevaluating the brain disease model of addiction

The Lancet Psychiatry, Год журнала: 2025, Номер unknown

Опубликована: Март 1, 2025

Язык: Английский

---

Machine learning-driven risk prediction and feature identification for major depressive disorder and its progression: an exploratory study based on five years of longitudinal data from the US national health survey

Journal of Affective Disorders, Год журнала: 2025, Номер unknown

Опубликована: Апрель 1, 2025

Язык: Английский

---

Detection of Suicidality Through Privacy-Preserving Large Language Models

medRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Март 8, 2024

Abstract Importance Attempts to use Artificial Intelligence (AI) in psychiatric disorders show moderate success, high-lighting the potential of incorporating information from clinical assessments improve models. The study focuses on using Large Language Models (LLMs) manage unstructured medical text, particularly for suicide risk detection care. Objective aims extract about suicidality status admission notes electronic health records (EHR) privacy-sensitive, locally hosted LLMs, specifically evaluating efficacy Llama-2 Main Outcomes and Measures compares performance several variants open source LLM extracting reports against a ground truth defined by human experts, assessing accuracy, sensitivity, specificity, F1 score across different prompting strategies. Results A German fine-tuned model showed highest accuracy (87.5%), sensitivity (83%) specificity (91.8%) identifying suicidality, with significant improvements various prompt designs. Conclusions Relevance demonstrates capability Llama-2, accurately while preserving data-privacy. This suggests their application surveillance systems emergencies improving management systematic quality control research. Key Points Question Can large language models (EHR)? Findings In this analysis 100 models, (Emgerman) demonstrated indicating models’ effectiveness on-site processing documentation detection. Meaning highlights records, data privacy. It recommends further these integrate them into enhance research mental

Язык: Английский

---