Depression clinical detection model based on social media: a federated deep learning approach

Abstract

Abstract Depression can significantly impact people’s mental health, and recent research shows that social media can provide decision-making support for healthcare professionals and serve as supplementary information for understanding patients’ health status. Deep learning models are also able to assess an individual’s likelihood of experiencing depression. However, data availability on social media is often limited due to privacy concerns, even though deep learning models benefit from having more data to analyze. To address this issue, this study proposes a methodological framework system for clinical decision support that uses federated deep learning (FDL) to identify individuals experiencing depression and provide intervention decisions for clinicians. The proposed framework involves evaluation of datasets from three social media platforms, and the experimental results demonstrate that our method achieves state-of-the-art results. The study aims to provide a personalized clinical decision support system with evolvable features that can deliver precise solutions and assist healthcare professionals in medical diagnosis. The proposed framework that incorporates social media data and deep learning models can provide valuable insights into patients’ health status, support personalized treatment decisions, and adapt to changing healthcare needs.