A Normative EMG Pipeline for Walking-Related Neuromuscular Coordination in Healthy Adults: A Foundation for Mild Cognitive Impairment Research

Abstract

Mild cognitive impairment (MCI) is associated with subtle motor abnormalities, yet current motor assessments often lack sensitivity and mechanistic specificity. This thesis developed and partly validated a Python-based electromyography (EMG) pipeline to quantify neuromuscular coordination during walking in healthy adults, providing a methodological foundation for future MCI studies. Seven cognitively unimpaired adults completed bilateral lower-limb EMG recordings during normal-speed treadmill, fast treadmill, and overground walking, and six returned for a repeat session to assess test–retest reliability. The pipeline integrated automated signal quality assessment, pre-processing, intermuscular coherence analysis, muscle activation sequence analysis, laterality analysis, and reliability testing. The knee muscle pair (L_VL↔L_BF) showed the strongest coherence and the most favourable reliability, especially during fast walking. Coherence varied across walking conditions, with overground walking associated with higher pipeline-specific beta-band coherence and fast walking revealing condition-dependent alpha-band changes. Activation sequence analysis demonstrated consistent spatiotemporal organisation across subjects, while rising slope metrics captured condition-related differences in activation dynamics. Reliability was acceptable for selected metrics but remained limited for others, particularly ankle and bilateral muscle pairs. These findings support the pipeline as a feasible and informative method for characterising walking-related neuromuscular coordination in healthy adults and provide preliminary normative reference data for future MCI research.