Study on the Relationship between Macroscopic Intrinsic Dynamics and Microcircuit Functions in Epilepsy

Research Background

Epilepsy is a group of neurological disorders characterized by abnormal spontaneous brain activity, involving multi-scale changes in brain functional organization. However, it remains unclear to what extent epilepsy-related spontaneous brain activity disturbances affect macroscopic intrinsic dynamics and microcircuit organization, and how these changes support its pathological relevance. Therefore, studying how spontaneous brain activity in epileptic patients affects macroscopic dynamics and microcircuit functions and exploring its pathological mechanisms has significant scientific importance.

Research Source

This research was jointly completed by the following scholars: Siqi Yang (School of Cyberspace Security, Chengdu University of Information Technology), Yimin Zhou (School of Cyberspace Security, Chengdu University of Information Technology), Chengzong Peng (School of Cyberspace Security, Chengdu University of Information Technology), Yao Meng (School of Life Science and Technology, University of Electronic Science and Technology of China), Huafu Chen (School of Life Science and Technology, University of Electronic Science and Technology of China), Shaoshi Zhang (Centre for Sleep and Cognition, Yong Loo Lin School of Medicine, National University of Singapore), Xiaolu Kong (Centre for Translational Magnetic Resonance Research, Yong Loo Lin School of Medicine, National University of Singapore), Ru Kong (Centre for Translational Magnetic Resonance Research, Yong Loo Lin School of Medicine, National University of Singapore), B. T. Thomas Yeo (Centre for Sleep and Cognition, Yong Loo Lin School of Medicine, National University of Singapore), Wei Liao (School of Life Science and Technology, University of Electronic Science and Technology of China), Zhiqiang Zhang (Neuroimaging Laboratory, Jinling Hospital, Medical School of Nanjing University). The article was published in the journal Communications Biology.

Research Methods

a) Research Procedure

The research is mainly divided into the following steps:

1. Selection of study subjects: Including 75 patients with Temporal Lobe Epilepsy (TLE), 79 patients with Genetic Generalized Clonic Seizures (GTCS), and 108 healthy controls (HC).
2. Time series feature extraction: Using fMRI BOLD time series to extract a large number of time features to characterize macroscopic intrinsic dynamics.
3. Macroscopic intrinsic dynamics analysis: Using Principal Component Analysis (PCA) to extract major spatial gradients from time series features.
4. Microcircuit simulation: Using the Parameter Mean Field Model (PMFM) to simulate microcircuit functions, including Recurrent Connections (RC) and External Inputs (I).
5. Data analysis and correlation study: Analyzing the spatial correlation between macroscopic intrinsic dynamics and microcircuit functions in epilepsy.

Time Series Feature Extraction and Principal Component Analysis

1. Extracting fMRI BOLD using Desikan-Killiany anatomical parcellation: Extracting time series from 68 cortical regions.
2. Feature extraction: Using the Highly Comparative Time Series Analysis (HCTSA) toolbox to extract over 7000 features from the time series of each region.
3. Principal Component Analysis: Capturing the spatial gradients of the time series feature matrix using PCA.

Microcircuit Simulation

1. Parameter Mean Field Model (PMFM): Used to simulate local synaptic properties, calculate group average functional connectivity (FC) and structural connectivity, and compute functional connectivity dynamics (FCD) for each participant.
2. Estimation of neural dynamic parameters: Setting RC, I, and σ parameters as linear combinations of the FC gradients and estimating these parameters by minimizing the divergence between the empirical and simulated FC and FCD matrices.

Major Results

Macroscopic Intrinsic Dynamics

1. Time Series Feature Gradient: The first two principal components (PC1 and PC2) extracted by PCA respectively captured the ventromedial-lateral gradient and sensory-associative gradient. In TLE and GTCS, time series feature gradients differed significantly from those in healthy control groups.
2. Regional and network changes: PC1 increased in the somatosensory motor and visual cortex areas in TLE and GTCS, while it decreased in the cingulate cortex. PC2 increased in the bilateral insula and anterior cingulate cortex in TLE, while it decreased in the visual cortex in GTCS.

Microcircuit Simulation

1. RC and I parameters: RC decreased in the sensory-motor regions and increased in the associative cortex regions for TLE and GTCS. The I parameter increased in the sensory-motor regions in TLE and across the entire cortex in GTCS. Random permutation tests indicated significant decreases in RC in the somatosensory motor and fusiform gyrus regions.

Multi-scale Brain Function Correlation

The study shows a strong negative correlation between the differences in macroscopic intrinsic dynamics gradients and microcircuit recurrent connection changes in TLE and GTCS, highlighting the interaction between dysfunctions at microscopic and macroscopic levels. The study also verifies the similarity between actual BOLD signal gradients and PMFM simulated BOLD signal gradients, indicating a close relationship between macroscopic intrinsic dynamics and microcircuit functions in epilepsy.

Research Conclusions and Value

1. Scientific Value: This study reveals the impact of abnormal neural activity in epilepsy on macroscopic and high-order networks, indicating systematic abnormalities in brain hierarchical organization.
2. Application Value: The approach connecting macroscopic and microscopic brain functions provides a deeper understanding of the neural mechanisms of epilepsy, offering new perspectives for epilepsy diagnosis and treatment.

Research Highlights

1. Important Findings: Significant changes in macroscopic time series feature gradients and microcircuit recurrent connections in epilepsy, especially in the somatosensory motor and default mode networks.
2. Novelty: The study employs novel methods through the extraction and analysis of multi-dimensional time series features, comprehensively characterizing spontaneous brain activity in epilepsy.

Other Valuable Information

1. Method Innovation: The HCTSA toolbox and PMFM model used in this study provide innovative methods for analyzing time series features and simulating neural dynamics.
2. Interdisciplinary Collaboration: The collaboration of experts from different fields helps integrate various technical means, enhancing the depth and breadth of the research.

Conclusion

This study reveals the characteristics of neural activity in epilepsy and the interaction between its macro and micro-functional dysfunctions by comprehensively utilizing time series analysis, neural dynamic simulation, and multi-scale brain function correlation analysis, providing new scientific basis for understanding and treating epilepsy.