A simulation study to identify most reliable corticomuscular connectivity estimators under continuous and bursting beta activity

Accurate corticomuscular connectivity estimation is essential for understanding the neural mechanisms underlying motor behaviors. However, the absence of ground truth in experimental data makes it difficult to determine which connectivity estimators most reliably capture shared cortical and muscular oscillatory components. Here, we developed a simulation framework extending a motoneuronal model to generate coupled EEG and EMG signals through a controlled connectivity model. Continuous and bursting beta-band activity were implemented as the only shared components at different ground truth connectivity levels. We systematically evaluated multiple classical cortico-muscular connectivity estimators and quantified their performance using comparative metrics across repeated simulations. Results identify corticomuscular coherence, multitaper coherence and phase locking value as the most accurate estimators to recover the imposed connectivity under different oscillatory regimes. Our findings provide methodological guidance for selecting connectivity measures and may facilitate more reliable investigations of brain–muscle interactions in both healthy and pathological conditions, supporting future neurorehabilitation strategies.