Scholarlink

Background and Objective <p>Cortico-muscular coherence (CMC) reflects the functional coupling between cortical neural oscillations and muscular activity, typically assessed using simultaneously recorded Electroencephalography (EEG) and Electromyography (EMG). Traditional coherence-based methods primarily capture linear, frequency-specific synchronisation, often overlooking nonlinear and scale-dependent dynamics inherent in neuromuscular control. This study proposes a framework to quantify CMC using nonlinear and information-theoretic descriptors, and investigates how cortico-muscular interaction varies with physical load during visually cued flexion-extension tasks.</p> Methods <p>EEG and EMG signals were recorded simultaneously from five healthy right-handed male participants (mean age 26 ± 3 years) performing arm lifts under nine load conditions (no load to 5&#xa0;kg). EEG was acquired from three cortical channels (C3, C4, Cz) at 256&#xa0;Hz; EMG was recorded from two right-arm muscle groups — Biceps Brachii (M1) and Flexor Carpi Radialis (M2) — and downsampled to 256&#xa0;Hz to match the EEG sampling rate. Correlations were computed across all six EEG–EMG channel pairs using three complementary descriptors — Multifractal Detrended Cross-Correlation Analysis (MF-DXA), Mutual Information (MI), and Joint Entropy (JEn) — yielding 18 features. Nine 60-second segments per subject, corresponding to the nine loading conditions, were extracted using visual-cue onset timestamps. Statistical significance was assessed using the Friedman test and Conover-Iman post hoc analysis (<i>p</i> ≤ 0.01).</p> Results <p>All 18 features showed statistically significant differences across loading conditions (<i>p</i> &lt; 0.01) for all five participants, demonstrating that MF-DXA, MI, and JEn are sensitive to load-dependent changes in corticomuscular coupling. This is a preliminary study conducted on a small cohort; the limited sample size (five participants, nine loading conditions) precludes machine-learning-based generalisation, and any classification-style analysis is reported strictly as a feasibility observation rather than a validated result.</p> Conclusion <p>This study introduces a preliminary, cost-effective, and non-invasive methodology for quantifying cortico-muscular connectivity using nonlinear and information-theoretic descriptors. The proposed framework shows statistically significant sensitivity to load-dependent neuromuscular changes; clinical validation with a larger cohort is required before diagnostic claims can be made. </p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Quantifying corticomuscular interactions during weightlifting: a joint EEG-EMG study using multifractal correlation and information-theoretic approaches

  • 期刊论文

  • 发表日期：

    2026年7月9日

查看全文
  • Debanjan Parbat,
  • Monisha Chakraborty

摘要

Background and Objective

Cortico-muscular coherence (CMC) reflects the functional coupling between cortical neural oscillations and muscular activity, typically assessed using simultaneously recorded Electroencephalography (EEG) and Electromyography (EMG). Traditional coherence-based methods primarily capture linear, frequency-specific synchronisation, often overlooking nonlinear and scale-dependent dynamics inherent in neuromuscular control. This study proposes a framework to quantify CMC using nonlinear and information-theoretic descriptors, and investigates how cortico-muscular interaction varies with physical load during visually cued flexion-extension tasks.

Methods

EEG and EMG signals were recorded simultaneously from five healthy right-handed male participants (mean age 26 ± 3 years) performing arm lifts under nine load conditions (no load to 5 kg). EEG was acquired from three cortical channels (C3, C4, Cz) at 256 Hz; EMG was recorded from two right-arm muscle groups — Biceps Brachii (M1) and Flexor Carpi Radialis (M2) — and downsampled to 256 Hz to match the EEG sampling rate. Correlations were computed across all six EEG–EMG channel pairs using three complementary descriptors — Multifractal Detrended Cross-Correlation Analysis (MF-DXA), Mutual Information (MI), and Joint Entropy (JEn) — yielding 18 features. Nine 60-second segments per subject, corresponding to the nine loading conditions, were extracted using visual-cue onset timestamps. Statistical significance was assessed using the Friedman test and Conover-Iman post hoc analysis (p ≤ 0.01).

Results

All 18 features showed statistically significant differences across loading conditions (p < 0.01) for all five participants, demonstrating that MF-DXA, MI, and JEn are sensitive to load-dependent changes in corticomuscular coupling. This is a preliminary study conducted on a small cohort; the limited sample size (five participants, nine loading conditions) precludes machine-learning-based generalisation, and any classification-style analysis is reported strictly as a feasibility observation rather than a validated result.

Conclusion

This study introduces a preliminary, cost-effective, and non-invasive methodology for quantifying cortico-muscular connectivity using nonlinear and information-theoretic descriptors. The proposed framework shows statistically significant sensitivity to load-dependent neuromuscular changes; clinical validation with a larger cohort is required before diagnostic claims can be made.