Quantitative Electrophysiological Neuromuscular Characterization

点击翻译按钮获取中文摘要

The structures and actions of nerves and muscles are studied by analyzing the electrophysiological signals they generate. Electromyographic (EMG) signals are used to study processes affecting skeletal muscle morphology and physiology. Skeletal muscles are organized into motor units (MUs); groups of fibres activated by single motor neurons. Each MU generates motor unit potentials (MUPs), which are summations of muscle fibre potentials (MFPs) generated by active MU fibres. During muscle contraction, MUs are repetitively active. Each active MU generates a motor unit potential train (MUPT), and an EMG signal is the summation of the MUPTs of active MUs. Neuromuscular disorders and aging affect the morphology and physiology of MUs, and the MFPs, MUPs and MUPTs they generate. Currently, the diagnosis of many neuromuscular disorders relies on subjective and qualitative analysis of the MUPs and MUPTs of needle-detected EMG signals. Patient outcomes are therefore highly dependent on individual physician expertise and training, and it is very difficult to assess disease progression and provide reliable prognoses. Current studies of the neuromuscular effects of aging use statistics of composite EMG signals, which cannot provide detailed individual MU information. As such, these methods are unable to detect important neuromuscular changes.***The proposed research objectives are to develop novel methods to create a quantitative electrophysiological neuromuscular characterization. These include methods to: 1) extract relevant MUP and MUPT data from EMG signals detected using standard methods; 2) suitably represent extracted MUP and MUPT data using sets of feature values which relate to specific aspects of MU morphology and physiology and; 3) apply sets of suitably represented MUPTs, relating to a representative set of MUs sampled from a test muscle to a classifier and interpretation module to provide a quantitative neuromuscular characterization. Novel signal processing and pattern recognition methods will be developed to extract MUP and MUPT data. Novel machine learning methods will be developed to interpret the data to provide decision support. The provided characterizations will be based on novel measures of MU fibre spatial arrangements, MFP temporal dispersion and MUP stability. This will allow them to be used to support and improve clinical decisions related to the diagnosis, treatment and management of neuromuscular disorders and provide better patient outcomes. The newly available detailed individual MU information will also facilitate the study of the neuromuscular effects of aging. Students working on these objectives will obtain valuable experience in both the natural sciences (neuromuscular physiology, electrophysiology) and engineering (signal processing, machine learning) and will be well equipped for employment related to the development of biomedical devices.