Related Experiment Video
Updated: May 26, 2026
08:43
Application of Granger Causality Analysis of the Directed Functional Connection in Alzheimer's Disease and Mild Cognitive Impairment
Published on: August 7, 2017
CCPD under sparsity and low-rank constraints: multi-frequency dynamic functional network connectivity analysis in
Li-Dan Kuang1, Yi-Wen Liu1, Ting Tang1
1School of Computer Science and Technology, Changsha University of Science and Technology, Changsha, Hunan, China.
Frontiers in Neuroscience
|May 25, 2026
Summary
Schizophrenia patients exhibit altered brain network dynamics, showing distinct connectivity patterns and temporal changes in functional networks compared to healthy controls. This research introduces a new computational method to identify these differences.
Area of Science:
- Neuroscience
- Computational Psychiatry
- Network Science
Background:
- Dynamic functional network connectivity (dFNC) analysis reveals brain network alterations in schizophrenia.
- Existing methods struggle to capture both shared patterns and specific temporal-frequency information in multi-frequency dFNC.
Purpose of the Study:
- To develop and apply a novel computational framework for jointly analyzing group-shared and group-specific features of multi-frequency dFNC in schizophrenia.
- To identify differences in dynamic functional connectivity between healthy controls and schizophrenia patients.
Main Methods:
- Utilized a coupled canonical polyadic decomposition (CCPD) approach, specifically a sparse and low-rank constrained CCPD (SLRCCPD) model.
- Decomposed multi-frequency dFNC tensors derived from the COBRE dataset (145 subjects: 71 SZs, 74 HCs).
- Incorporated L1-norm regularization and nuclear norm-based low-rank approximation for enhanced analysis.
Main Results:
- Identified significant connectivity differences between schizophrenia patients (SZs) and healthy controls (HCs) in auditory, somatomotor, cognitive control, visual, and cerebellar networks across five dynamic modules.
- SZs demonstrated significantly higher fractional and dwell time in a specific dynamic state (State 3) across low- and high-frequency bands.
- SZs exhibited fewer state transitions compared to HCs across all frequency bands.
Conclusions:
- The SLRCCPD framework effectively captures abnormal multi-band dynamic functional connectivity in schizophrenia.
- This approach provides a novel computational tool for investigating brain network dynamics in psychiatric disorders.
- Findings offer insights into the mechanistic underpinnings of schizophrenia through altered brain network temporal and frequency characteristics.
