نهمین کنگره نقشه برداری مغز

نمایش مقاله

Title:

Functional Cortico-Muscular Connectivity as a Quantitative Brain Mapping Framework: An Advanced EEG–EMG and Machine Learning Study

Subject:

Artificial Intelligence and Brain Mapping

Presenter Author:
Name:	Sajjad Abdollahi

Author:
Name 1:	Sajjad Abdollahi - Corresponding - Presenter
Affiliation 1:	PhD candidate, Department of Sport Biomechanics, Faculty of Sports Sciences, Bu-Ali Sina University, Hamedan, Iran

Name 2:	Mehrdad Anbarian
Affiliation 2:	Professor, Department of Sport Biomechanics, Faculty of Sports Sciences, Bu-Ali Sina University, Hamedan, Iran

Name 3:	Rahman Sheikhhoseini
Affiliation 3:	Associate Professor, Department of Corrective Exercise and Sport Injury, Faculty of Physical Education and Sport Sciences, Allameh Tabataba’i University, Tehran, Iran

Name 4:
Affiliation 4:

Name 5:
Affiliation 5:

Name 6:
Affiliation 6:

Name 7:
Affiliation 7:

Name 8:
Affiliation 8:

Name 9:
Affiliation 9:

Name 10:
Affiliation 10:

Abstract

Original Research

introduction :

Accurate brain mapping of motor control necessitates biomarkers that extend beyond isolated cortical activity and capture the dynamic interaction between the central nervous system and the musculoskeletal system. Functional cortico-muscular connectivity (CMC) provides a physiologically grounded measure of neural synchronization between motor cortical oscillations and peripheral muscle activation. The objective of this study was to establish CMC as a quantitative brain mapping framework by integrating high-density EEG–EMG recordings with advanced machine learning architectures during controlled static motor tasks.

Methods:

Simultaneous EEG and surface EMG data were collected from 24 physically active individuals performing standardized isometric force production tasks. EEG signals were acquired using a 64-channel active electrode system targeting primary and premotor cortices, while EMG activity was recorded from the quadriceps femoris, hamstring, and medial and lateral gastrocnemius muscles. Signal preprocessing included band-pass filtering (EEG: 1–45 Hz; EMG: 20–450 Hz), independent component analysis for artifact removal, and temporal normalization. CMC was quantified using magnitude-squared coherence within beta (13–30 Hz) and gamma (30–45 Hz) frequency bands. Connectivity features were integrated with biomechanical indices of force variability and stability. Supervised learning models—including support vector machines, random forest classifiers, multilayer perceptrons, and long short-term memory (LSTM) networks—were trained using a stratified 10-fold cross-validation protocol.

Results:

The machine learning–driven analysis revealed highly discriminative cortico-muscular connectivity patterns associated with motor control performance. Among the evaluated models, the LSTM architecture achieved superior classification accuracy (91.3%), outperforming SVM (87.4%), random forest (85.9%), and MLP (88.1%) models. Feature contribution analysis demonstrated that beta-band CMC accounted for approximately 62% of model explanatory power, underscoring its central role in stable force regulation. A strong positive association was observed between CMC magnitude and force stability indices (r = 0.71, p < 0.001), indicating that reduced neural synchronization corresponds to diminished motor control efficiency. Low-dimensional embedding of feature space revealed distinct and reproducible neural–muscular coordination signatures across task conditions.

Conclusion:

This study positions functional cortico-muscular connectivity as a robust and physiologically interpretable framework for brain mapping of motor control. By coupling EEG–EMG connectivity metrics with deep learning models, the proposed approach enables high-resolution characterization of neural–muscular interactions beyond conventional cortical mapping. The findings highlight the transformative role of artificial intelligence in advancing brain mapping methodologies and provide a scalable foundation for applications in clinical neurorehabilitation, motor performance optimization, and precision-based intervention design.

Keywords:

Brain Mapping, Electroencephalography, Electromyography, Functional Connectivity, Machine Learning, Motor Cortex