Identifying muscle activity patterns in static and dynamic activities of daily living

Human motor control is primarily studied inside labs or clinics, through analysis of predefined, constrained motor tasks. In these studies subjects conduct experimental tasks with biomechanical constraints, in ways potentially unrealistic for the individual. We propose that to better understand the mechanisms of motor control, subjects and tasks need to be analysed in daily life unconstrained scenarios, and produce tools to highlight the inter-subject variability and tailor the analysis to the uniqueness of the individual. Shared motor control strategies across populations require investigation but importance should be equally given to the individuality and personal muscle recruitment strategies. Muscle activity patterns (MAPs) were identified in three motor experiments based on daily living (ADLs) tasks, assessing static and dynamic scenarios. Using non-invasive methods, algorithms were designed to identify MAPs from electromyography (EMG) recordings in each ADLs, establishing relationships with recorded biomechanics and generated forces. Active structures of the central nervous systems (CNS) for motor coordination, were predicted through spectral analysis of the EMG signals. These findings highlight inter-subject variability and need for individualised assessment of motor control strategies, along with shared features in the MAPs (Article 3). Active neural pathways were teased for healthy individuals, exhibiting an oscillatory behaviour throughout the tasks (Articles 3 and 4). To address the need for ADLs recordings, a novel method is presented for sEMG segmentation, in unconstrained gait scenarios, in Article 4. Lastly, results of a human-robot-interaction study link MAPs with biomechanics to provide a biomarker for personalised optimal muscle recruitment, minimising effort (Article 5). The overall study comprises a series of novel tools, able to characterise motor tasks in terms of MAPs, biomechanics and generated forces, with higher versatility for studies outside of the lab or clinic. These findings may greatly improve current motor control diagnosis by providing quantitative measurements and biomarkers for individually-tailored neurorehabilitation.