Related Experiment Video
Updated: Aug 23, 2025
05:19
Author Spotlight: Therapeutic Benefit of Closed-Loop Deep Brain Stimulation in Depression Treatment
Published on: July 7, 2023
2.4K
Decision support system for the differentiation of schizophrenia and mood disorders using multiple deep learning
Duc-Khanh Nguyen1, Chien-Lung Chan2, Ai-Hsien A Li3
1Department of Information Management, 34895Yuan Ze University, Taoyuan, Taiwan.
Health Informatics Journal
|November 1, 2022
Summary
This study introduces a deep learning (DL) system using wearable device data to help differentiate mental health disorders (MHDs). The AI model shows promise as an objective tool for physicians in diagnosing conditions like schizophrenia and mood disorders.
Area of Science:
- Computational psychiatry
- Digital health
- Machine learning in medicine
Background:
- Mental health disorders (MHDs) are increasingly prevalent globally, posing significant societal and familial burdens.
- Differentiating between various MHDs is challenging due to overlapping symptoms and varying severity.
- Objective diagnostic tools are needed to support clinical decision-making in mental healthcare.
Purpose of the Study:
- To develop and evaluate a deep learning (DL) based support system for the objective differentiation of mental health disorders.
- To utilize data from wearable devices to aid physicians in differential diagnosis and treatment planning for MHDs.
Main Methods:
- Experiments were conducted using open datasets (Psykose and Depresjon) containing activity motion signal data from wearable devices.
- A deep learning framework was developed to analyze motion data for identifying specific MHDs, including schizophrenia and mood disorders (bipolar and unipolar).
- The proposed DL framework's performance was compared against traditional machine learning (ML) and other DL methods.
Main Results:
- The proposed DL framework demonstrated strong performance in differentiating between schizophrenia and mood disorders using wearable sensor data.
- Comparative analysis indicated that the DL approach outperformed traditional ML methods in this diagnostic support task.
- Both workflow approaches within the proposed framework yielded favorable results, highlighting its robustness.
Conclusions:
- Deep learning models applied to activity motion signal data from wearable devices offer a prospective objective support system for MHD differentiation.
- The system shows good performance, suggesting its potential utility in clinical settings for enhancing diagnostic accuracy.
- This approach represents a significant advancement in leveraging digital health technologies for mental health assessment.
