Dataset for Systematic Review: Computational Strategies for Depression Detection and Treatment

Tallón, Paula

doi:10.5281/zenodo.17978674

Dataset for Systematic Review: Computational Strategies for Depression Detection and Treatment

Tallón, Paula

Publisher: Zenodo
DOI: 10.5281/zenodo.17978674
Repository: zenodo.org
Language: en

Abstract

Systematic review metadata and supplementary materials for the study: Computational Strategies for Depression Detection and Treatment: The Role of Behavioral Activation and Neurobiological Insights – A Systematic Review This dataset contains structured metadata for the 59 studies included in the PRISMA 2020-compliant systematic review. It supports the analysis of artificial intelligence (AI), machine learning, behavioral activation (BA), physical activity monitoring, and neurobiological mechanisms in depression detection and treatment. Contents:- `dataset.csv`: Full bibliographic details (Study ID, first author, year, full title, journal, AI technique, accuracy/outcome metrics, population, DOI/URL) for all 59 included studies, plus 2 more regarding the references of the checklist of CLAIM and IJMI AI/ML reporting guidelines.- `S1_File.docx`: Retrospective review protocol, including PICO framework, search strategy, and inclusion/exclusion criteria.- `S1_Table.pdf`: Complete search strings used across PubMed, Scopus, ACM Digital Library, and Web of Science.- `S2_File.docx`: Completed PRISMA 2020 Checklist.- `README.markdown`: Dataset overview and usage instructions.- `data_availability_statement.md`: Final data availability statement for inclusion in the manuscript.Keywords: depression, artificial intelligence, behavioral activation, machine learning, EEG, neuroimaging, digital mental health, PRISMA-'Appendix_S2_AI_ML_reporting_checklist.docx': Compliance with CLAIM and IJMI AI/ML reporting guidelines

Full text

S1 Table: Search Terms and Strategy This table lists the search strings, variants, and additional terms used across six databases to identify studies for the systematic review. Database Search String Variants/Additional Terms Notes PubMed depression AND (behavioral activation OR physical activity OR frailty OR neural networks OR brain phenotype OR computational) Depression: major depressive disorder, MDD, depressive symptoms, dysthymia Behavioral Activation: BA, activity scheduling Physical Activity: exercise, motor activity Frailty: sarcopenia, physical decline Neural Networks: deep learning, machine learning, AI, artificial intelligence Brain Phenotype: neuroimaging, EEG, fMRI, brain imaging, neurobiology Computational: bioinformatics, data science, algorithms MeSH terms applied (e.g., “Depressive Disorder”[Mesh]). Filters: Peer-reviewed, 2010–2025. Scopus TITLE-ABS-KEY(depression AND (behavioral activation OR physical activity OR frailty OR neural networks OR brain phenotype OR computational)) Same as PubMed, plus: Depression: mood disorders Neural Networks: convolutional neural networks, CNN, RNN Brain Phenotype: structural MRI, functional connectivity Limited to articles, reviews, 2010–2025. ACM depression AND (behavioral activation OR physical activity OR frailty OR neural networks OR brain phenotype OR computational) Focus on computational terms: machine learning, AI models, predictive modeling, data mining Neural Networks: supervised learning, unsupervised learning Targeted computer science literature. Web of Science TS=(depression AND (behavioral activation OR physical activity OR frailty OR neural networks OR brain phenotype OR computational)) Same as PubMed/Scopus, plus: Brain Phenotype: cortical thickness, white matter integrity Refined by Web of Science categories (Neurosciences, Psychiatry, Computer Science). PsycINFO depression AND (behavioral activation OR physical activity OR frailty OR neural networks OR brain phenotype OR computational) Same as PubMed, plus: Depression: mood disorders Behavioral Activation: activity engagement Limited to peer-reviewed, 2010–2025. IEEE Xplore depression AND (behavioral activation OR physical activity OR frailty OR neural networks OR brain phenotype OR computational) Focus on computational terms: AI, machine learning, neural networks Brain Phenotype: EEG signal processing Targeted engineering literature.