The neural modeling of sympathetic nerve activity and baroreflex hysteresis for the regulation of arterial pressure in humans

This dissertation analyzes two important mechanisms within the integrated neural system that regulates blood pressure---the baroreflex. The first part introduces new signal restoration algorithms based upon a neural firing model as alternatives to standard methods to preprocess efferent sympathetic nervous activity (SNA). Conventional protocols filter, full-wave rectify, threshold, and integrate the grouped nerve impulses to generate a neurogram of simple bursts of nerve activity. The current analysis proposes a sympathetic nerve firing model that requires the specification of an action potential template. Signal and noise analyses show that nerve spikes can be extracted based upon amplitude and firing time. The spike template is incorporated into a new restoration algorithm based on l1-regularization (Donoho et al., 1992). The restored neurogram is compared to the conventional method via a validation model to predict SNA from lagged blood pressure. The results show that application of both the denoising of raw SNA and l 1-regularization restores the neurogram significantly better than the conventional protocols.; The second part of the dissertation models a phenomenon called baroreflex hysteresis, characterized by control of nervous outflow that is dependent upon the direction of afferent baroreceptor input. This pattern of neural outflow is influenced by the two primary system components---mechanics of the vessels with embedded baroreceptor afferent nerves and responsiveness of the afferent-efferent neural pathway. The state-dependent relationship among pressure as the initial system input, vessel diameter as the transducer of pressure, and heart rate as the result of afferent-efferent neural outflow is linearly modeled as a three-dimensional planar ellipse whose dynamics can be characterized by two harmonic oscillators. Two-dimensional projections of this ellipse provide direction of motion and quantitate hysteresis in the mechanical (pressure---diameter), neural (diameter---heart rate), and integrated baroreflex (pressure---heart rate) components. Baroreflex hysteresis can be modeled as the linear combination of hysteresis in both vessel mechanics and afferent-efferent neural control weighted by their directions of motion. This model of baroreflex system allows assessment of the relative importance of neural versus non-neural contributions to baroreflex responses.