Artificial Intelligence for Early Detection of Anxiety and Depression: A Rapid Review and Theoretical Perspective on Speech, Text, and Online Behavioral Indicators

Document Type : Original Article

Authors

1 Payam-e Noor University, Isfahan (Mobarakeh Center) Branch, Isfahan, Iran.

2 Department of Counselling, Faculty of Educational Sciences and Psychology, University of Isfahan, Isfahan, Iran.

3 Department of Clinical Psychology, Islamic Azad University, Isfahan (Khorasgan) Branch, Isfahan, Iran.

Abstract

Anxiety and depression are among the most prevalent mental health disorders worldwide, and early detection remains a clinical priority. This rapid review synthesizes evidence on the application of artificial intelligence (AI) for detecting anxiety and depression using speech, text, and online behavioral data. A systematic search was conducted in Scopus, PubMed, IEEE Xplore, ACM Digital Library, Web of Science, and Google Scholar for English-language studies published between 2013 and 2025. Studies applying machine learning, deep learning, natural language processing (NLP), or speech analysis for mental health detection were included. Fifteen studies (including empirical, review, and model-based studies) met the inclusion criteria. AI models have been shown to identify depression-related indicators, such as negative language, self-referential expressions, reduced social interaction, and acoustic speech changes. Anxiety-related indicators included worry-related language, emotional instability, and irregular behavioral patterns. Multimodal and transformer-based models demonstrated improved contextual understanding. AI-based systems show potential for early detection of anxiety and depression; however, they should not be considered diagnostic tools due to limitations in data quality, bias, and clinical validation.

Keywords

Main Subjects

A Review of Theorizing of Behavioral Sciences
Volume 2, Issue 4
October 2025
Pages 1-14

Files

Share

How to cite

Statistics