Multiscale Intermuscular Coupling Analysis via Complex Network-Based High-Order O-Information.

Area of Science:

• Biomechanics
• Neuroscience
• Complex Systems Analysis

Background:

• Intermuscular coupling analysis (IMC) is crucial for understanding muscle control and stroke rehabilitation.
• Current understanding of IMC lacks insight into higher-order interactions and microscopic characteristics.

Purpose of the Study:

• To develop a multiscale framework for IMC using complex networks and O-Information.
• To analyze microscopic neural information by decomposing surface electromyography (sEMG) signals into motor units (MU).
• To investigate higher-order interactions and network characteristics in human upper limb movements.

Main Methods:

• Developed a multiscale intermuscular coupling analysis framework incorporating complex networks and O-Information.
• Decomposed sEMG signals to obtain motor units (MU) for microscopic analysis.
• Applied graph theory and O-Information analysis to data from three upper limb movements.

Main Results:

• Graph analysis revealed significant differences in muscle network connectivity across movement tasks.
• Motor unit analysis showed a mismatch between muscle distribution and motor neuron inputs, indicating reduced control dimensionality during multi-joint tasks.
• O-Information analysis identified prevalent redundant information in sEMG and MU networks, alongside low-order synergistic subsystems.

Conclusions:

• The proposed framework offers a comprehensive approach to IMC, integrating macroscopic and microscopic network features.
• The findings highlight potential for engineering applications, demonstrated by promising classification accuracy using K-Nearest Neighbors (KNN).
• This research advances the understanding of muscle coordination and neural control mechanisms.