Adaptive change point detection of dynamic functional connectivity networks

Area of Science:

• Neuroscience
• Medical Imaging
• Signal Processing

Background:

• Resting-state functional magnetic resonance imaging (rsfMRI) is crucial for studying intrinsic brain functionality.
• Brain functional networks dynamically change states, and identifying these change points is vital for understanding normal and disordered brain function.
• Detecting these change points is challenging due to unknown underlying factors and the absence of ground truth.

Purpose of the Study:

• To develop and validate a novel algorithm for adaptively detecting change points in brain functional connectivity networks.
• To enhance the understanding of dynamic brain network alterations in various conditions.
• To provide a reliable method for identifying critical transitions in brain activity.

Main Methods:

• The study utilizes resting-state functional magnetic resonance imaging (rsfMRI) data.
• A new adaptive algorithm is proposed, employing the sum of absolute sign differences between adjacent images from rsfMRI scans.
• Measures from image and video processing are integrated into the algorithm for change point detection.

Main Results:

• The proposed algorithm effectively detects change points in functional connectivity networks.
• Reliable detection of change points was demonstrated in both task-based and resting-state brain networks.
• The findings highlight the algorithm's robustness and potential for clinical applications.

Conclusions:

• The developed adaptive algorithm offers a reliable method for detecting dynamic changes in brain functional networks.
• Understanding these change points can significantly aid in the identification and characterization of brain disorders.
• The study opens new avenues for future research in neuroimaging and brain connectivity analysis.