Quantitative Identification of Major Depression Based on Resting-State Dynamic Functional Connectivity: A Machine...

Area of Science:

• Neuroscience
• Medical Imaging
• Machine Learning

Background:

• Accurate diagnosis of major depressive disorder (MDD) is crucial for effective intervention.
• Traditional machine learning methods using static functional connectivity (SFC) show limitations in identifying MDD.
• Exploiting dynamic functional connectivity (DFC) offers a promising avenue for improved MDD detection.

Purpose of the Study:

• To develop an accurate and objective image-based diagnostic system for MDD.
• To compare the diagnostic performance of DFC versus SFC in identifying MDD patients.
• To investigate the spatiotemporal characteristics of brain connectivity differentiating MDD from healthy controls.

Main Methods:

• Collected MRI data from 99 participants (56 healthy controls, 43 MDD patients).
• Calculated DFC using a sliding-window algorithm and employed a non-linear support vector machine (SVM) classifier.
• Analyzed the spatiotemporal features of discriminative brain connections.

Main Results:

• The SVM classifier achieved an AUC of 0.9913 using DFC, significantly higher than 0.8685 using SFC.
• Discriminative connections were identified across multiple brain networks, including the frontoparietal network (FPN), default mode network (DMN), and visual network (VN).
• Temporally, key connections showed a transition from cortical to deeper brain regions.

Conclusions:

• DFC provides a superior and reliable quantitative method for identifying MDD compared to SFC.
• The findings enhance understanding of the neural mechanisms underlying MDD.
• This approach holds potential for improving the accuracy of MDD diagnosis and facilitating early intervention.