Convolutional Neural Network for Depression and Schizophrenia Detection.

Area of Science:

• Artificial Intelligence
• Computer Vision
• Mental Health Diagnostics

Background:

• Mental health disorders like depression and schizophrenia can manifest in subtle motor activity changes.
• Traditional diagnostic methods may involve subjective self-reporting or invasive procedures.

Purpose of the Study:

• To develop a Convolutional Neural Network (CNN) model for detecting depression and schizophrenia.
• To utilize motor activity patterns, visualized as images, for mental disorder classification.
• To create a non-invasive support tool for mental health diagnosis.

Main Methods:

• Motor activity data were captured and transformed into image representations.
• A CNN model was trained and tested on these visual motor patterns.
• Three-fold cross-validation was employed to evaluate model performance.

Main Results:

• The CNN model achieved an average accuracy of 95% in classifying mental health conditions.
• CNNs successfully identified discriminative motor activity features differentiating between patient groups.
• The approach demonstrated effectiveness in distinguishing individuals with depression, schizophrenia, and healthy controls.

Conclusions:

• Computer vision and deep neural networks show significant potential for early, non-invasive mental health diagnosis.
• Motor activity analysis via CNNs can serve as a valuable tool in mental health support.
• This method offers a promising avenue for objective assessment of psychological conditions.