Analyzing mental disorders with a CNN-GRU deep learning model on motor activity.

Area of Science:

• Neuroscience
• Computer Science
• Psychiatry

Background:

• Mood disorders significantly impair daily functioning and well-being.
• Global mental health resources are scarce, and stigma remains a barrier.
• Current mood disorder detection methods rely on static data and are often invasive.

Purpose of the Study:

• To explore the potential of continuous motor activity data for mood disorder detection.
• To develop and evaluate a Deep Learning Model for analyzing motor activity sequences.
• To provide an objective, non-invasive method for ongoing mood state monitoring.

Main Methods:

• Utilized a Convolutional Neural Network-Gated Recurrent Unit (CNN-GRU) deep learning architecture.
• Analyzed continuous motor activity sequences from wrist-worn actigraphy data.
• Experimented on depression datasets to assess model performance.

Main Results:

• The CNN-GRU model achieved a high accuracy of 98.1% in detecting mood disorders.
• This performance surpasses existing state-of-the-art techniques.
• Continuous motor activity analysis proved effective for mood disorder assessment.

Conclusions:

• Deep learning analysis of motor activity data is a viable and accurate method for mood disorder detection.
• Wearable actigraphy offers an objective, non-invasive approach for continuous mental health monitoring.
• This technology has the potential to improve accessibility and reduce stigma in mental healthcare.