Wed, Nov 11, 2026
| فارسی
Volume 32, Issue 1 (Continuously Updated 2026)
IJPCP 2026, 32(1): 0-0 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Bagheri Sheykhangafshe F, Tajbakhsh K, Savabi Niri V, Ramezanpour Mermeti F, Afsarpour S A. Transforming Suicide Risk Detection and Prevention through Artificial Intelligence and Machine Learning: A Systematic Review. IJPCP 2026; 32 (1)
URL: http://ijpcp.iums.ac.ir/article-1-4598-en.html
URL: http://ijpcp.iums.ac.ir/article-1-4598-en.html
Farzin Bagheri Sheykhangafshe1 
, Khazar Tajbakhsh2 , Vahid Savabi Niri3 , Fatemeh Ramezanpour Mermeti4 , Seyedeh Atousa Afsarpour5
, Khazar Tajbakhsh2 , Vahid Savabi Niri3 , Fatemeh Ramezanpour Mermeti4 , Seyedeh Atousa Afsarpour5
1- Department of Psychology, Faculty of Humanities, Tarbiat Modares University, Tehran, Iran. , farzinbagheri@modares.ac.ir
2- Department of Psychology, University of Milano-Bicocca, Milan, Italy.
3- Department of Clinical Psychology, Ard.C., Islamic Azad University, Ardabil, Iran.
4- Department of Clinical Psychology, Sar.C., Islamic Azad University, Sari, Iran.
5- Department of Positive Islamic Psychology, Behshahr Branch, Payame Noor University, Behshahr, Iran.
2- Department of Psychology, University of Milano-Bicocca, Milan, Italy.
3- Department of Clinical Psychology, Ard.C., Islamic Azad University, Ardabil, Iran.
4- Department of Clinical Psychology, Sar.C., Islamic Azad University, Sari, Iran.
5- Department of Positive Islamic Psychology, Behshahr Branch, Payame Noor University, Behshahr, Iran.
Abstract: (535 Views)
Objectives Given the rapid growth in the use of digital and computerized tools in mental health, there is an increasing need to systematically review the current evidence on the use of advanced technologies. This review aimed to review studies that applied artificial intelligence (AI) and machine learning (ML) algorithms for suicide risk identification and prevention.
Methods This is a systematic review in accordance with the PRISMA guidelines. We retrieved relevant articles in English published from January 2014 to September 2025 from Google Scholar, PubMed, ProQuest, EMBASE, PsycINFO, Web of Science, CINAHL, and Scopus. Initially, 751 articles were identified. Following a comprehensive quality assessment using a standardized checklist, 25 studies with the highest methodological quality were selected for the review.
Results The studies utilized diverse data sources, including text messages, social media platforms, electronic health records, and psychiatric clinical notes, with sample sizes ranging from a few hundred to several million records. The applied AI and ML approaches included logistic regression, random forests, decision trees, XGBoost, and deep neural networks, with prediction accuracies ranging from 65% to over 95%. Key risk factors identified across studies included psychiatric disorders, depression, method of suicide attempt, marital status, low education, unemployment, and social isolation. Random forest and deep neural networks consistently demonstrated the highest predictive performance.
Conclusion AI and ML algorithms can detect complex patterns of suicide risk based on both clinical and social media data, enabling early prediction and timely intervention. Random forest and deep neural network models have the highest accuracy and are practical tools for designing effective preventive strategies.
Methods This is a systematic review in accordance with the PRISMA guidelines. We retrieved relevant articles in English published from January 2014 to September 2025 from Google Scholar, PubMed, ProQuest, EMBASE, PsycINFO, Web of Science, CINAHL, and Scopus. Initially, 751 articles were identified. Following a comprehensive quality assessment using a standardized checklist, 25 studies with the highest methodological quality were selected for the review.
Results The studies utilized diverse data sources, including text messages, social media platforms, electronic health records, and psychiatric clinical notes, with sample sizes ranging from a few hundred to several million records. The applied AI and ML approaches included logistic regression, random forests, decision trees, XGBoost, and deep neural networks, with prediction accuracies ranging from 65% to over 95%. Key risk factors identified across studies included psychiatric disorders, depression, method of suicide attempt, marital status, low education, unemployment, and social isolation. Random forest and deep neural networks consistently demonstrated the highest predictive performance.
Conclusion AI and ML algorithms can detect complex patterns of suicide risk based on both clinical and social media data, enabling early prediction and timely intervention. Random forest and deep neural network models have the highest accuracy and are practical tools for designing effective preventive strategies.
Type of Study: Rewie |
Subject:
Psychiatry and Psychology
Received: 2025/08/30 | Accepted: 2026/01/10 | Published: 2026/02/23
Received: 2025/08/30 | Accepted: 2026/01/10 | Published: 2026/02/23
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
Copyright © The Author(s);
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC-By-NC), which permits use, distribution, and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.
Contact Information
