Leveraging Deep Learning for Depression Diagnosis in Women: A Hybrid CNN-GRU-MLP Fusion Model

The prevalence of mental health disorders, particularly depression, has significantly increased among women in recent years, necessitating early and accurate detection for timely intervention.Traditional diagnostic methods, such as self-reported surveys and clinical assessments, often suffer from subjectivity, inconsistencies, and accessibility issues.Existing machine learning-based approaches, which predominantly rely on standalone deep learning models, face challenges in effectively capturing both spatial and temporal dependencies in mental health data.To address these limitations, this study introduces a novel hybrid model that integrates Convolutional Neural Networks (CNN), Gated Recurrent Units (GRU), and Multi-Layer Perceptron (MLP) to enhance depression classification accuracy.The CNN component extracts spatial feature patterns, GRU captures temporal dependencies in sequential data, and MLP improves high-dimensional feature learning, leading to more robust and reliable predictions.The model is trained on a dataset collected through a nationwide survey of women experiencing psychological distress, which has been curated and made publicly available on the Kaggle repository for further research and validation.This dataset comprises 21 distinct features, along with a ground truth label indicating clinical diagnosis.Performance evaluations demonstrate that the proposed fusion model outperforms traditional approaches, achieving an accuracy of 97.8861%, recall of 97.8861%, precision of 97.887%, and an F1-score of 97.885%.The results indicate a significant improvement over existing methods, providing a scalable, automated, and highly accurate tool for depression detection in women.This study contributes to mental health research by offering an open-access dataset and a robust deep learning model, fostering further advancements in depression detection.