Reproduce “A Comprehensive Review of Datasets for Clinical Mental Health AI Systems” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Aishik Mandal, Prottay Kumar Adhikary, Hiba Arnaout, Iryna Gurevych, Tanmoy Chakraborty
A review systematically categorizes 89 clinical mental health datasets, examining their modalities, accessibility, disorder focus, and sociocultural context to inform the development of AI-powered clinical assistants. The analysis identifies critical gaps in data diversity, standardization, and multimodal synthetic generation, proposing strategies for more robust and equitable AI systems.
Key insights
- The majority of clinical mental health datasets are private or restricted, severely limiting data sharing, reproducibility, and generalizability for AI research.
- Dataset coverage is uneven, with schizophrenia, PTSD, and depression being the most represented disorders, and a significant lack of data for anxiety and bipolar disorder.
- Current synthetic data generation largely focuses on text-based modalities and single-turn interactions, with a substantial gap in multimodal (speech, video) and longitudinal synthetic data.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.