A statistical framework for the analysis of neural control of movement with aging and other clinical applications
Johnson, Ashley Nzinga
MetadataShow full item record
The majority of daily living tasks necessitate the use of bimanual movements or concurrent cognitive processing, which are often more difficult for elderly adults. With the number of Americans age 65 and older expected to double in the next 25 years, in-depth research and sophisticated technologies are necessary to understand the mechanisms involved in normal neuromuscular aging. The objective of the research is to understand the effects of aging on biological signals for motor control and to develop a methodology to classify aging and stroke populations. The methodological approach investigated the influence on correlated activity (coherence) between electroencephalogram (EEG) and electromyogram (EMG) signals into senior age. In support of classifying aging and stroke populations, the methodology selected optimal features from the time, frequency, and information theory domains. Additionally, the use of cepstral analysis was modified toward this application to analyze EEG and EMG signals. The inclusion and optimization of cepstral features significantly improved classification accuracy. Additionally, classification of young and elderly adults using Gaussian Mixture Models with Minimum Classification Error improved overall accuracy values. Contributions from the dissertation include demonstration of the change in correlated activity between EMG and EEG with fine motor simple and complex dual tasks; a quantitative feature library for characterizing the neural control of movement with aging under three task conditions; and a methodology for the selection and classification of features to characterize the neural control of movement. Additionally, the dissertation provides functional insight for the association of features with tasks, aging, and clinical conditions. The results of the work are significant because classification of the neural control of movement with aging is not well established. From these contributions, future potential contributions are: a methodology for physiologists to analyze and interpret data; and a computational tool to provide early detection of neuromuscular disorders.