On the use of information theory for the analysis of the neurophysiological control of muscle force

• Grant number: 19/01508-4
• Support type: Scholarships in Brazil - Doctorate
• Effective date (Start): May 01, 2019
• Effective date (End): April 30, 2022
• Field of knowledge: Engineering - Biomedical Engineering
• Principal Investigator: Leonardo Abdala Elias
• Grantee: Ellen Pereira Zambalde
• Home Institution: Faculdade de Engenharia Elétrica e de Computação (FEEC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil

Abstract

Studies on the neurophysiological control of muscle force have used correlation analysis to understand the functional coupling between different structures of the nervous system during a motor task. The assumption behind these techniques is that the relations between the experimentally recorded variables [e.g., electroencephalogram (EEG), surface electromyogram (sEMG), motor unit activity] are linear. However, the assumption on linearity might not be strictly valid for the neuromuscular system, even during the maintenance of an approximately steady isometric contraction. Therefore, investigations on the applicability of new techniques for the analysis of neural signals produced during force control are necessary. In this vein, the use of metrics based on information theory would be an appropriate solution to the problem, since these metrics are model independent, and can capture both linear and nonlinear interactions between variables. In this doctoral research project, metrics based on information theory (e.g., entropy and mutual information) will be evaluated for the study of the neurophysiological control of muscle force. Experiments will be carried out on healthy participants during the performance of isometric contractions with different intensities. Force, EEG, and the activities of populations of motor units extracted from high-density sEMG will be recorded in the experiments. Computer simulations of a multiscale mathematical model of the neuromuscular system will be performed to evaluate the factors influencing information theory metrics. The metrics based on information theory and the traditional metrics based on correlation analysis will be compared to evaluate the limits of application of information theory to understand the functioning of the neuromuscular system. (AU)