Morales, George J.

Coherence estimates have been used to determine the presence of functional coupling between two signals. While direct projections from the nucleus reuniens (RE) to the hippocampus formation in the rat have been discovered, little is known about the possible functional influence of the RE on the hippocampus. This investigation makes use of MATLAB to create a set of specialized algorithms to investigate coherence function estimates between RE cell activity and hippocampal EEG. In addition, error prevention considerations as well as shortcomings in current data acquisition software that ultimately lead to the necessity for additional software analysis tools are also discussed. An investigation into RE cell behavior requires the calculation of cell activity spike rates as well as the identification of action potential bursting phenomena. Isolation of individual cell activity, from a population recording channel, is needed in order to prevent erroneous effects associated with using unresolved multi-neuron recordings. Changes in spike rate activity and frequency of bursting occurrences are calculated as a means of gauging RE unit response to the presence of a stimulus (e.g., tail pinch). The relationship of RE units on hippocampal EEG by analysis of coherence function estimates between RE units and hippocampal EEG, as well as evaluated RE unit behavior in terms of changes in unit spike rate and bursting activity are established.

A method for modeling and simulating neural action potential (AP) propagation along the length of an axon containing a number of Ranvier nodes is proposed in this dissertation. A system identification approach is adopted to represent node of Ranvier (NR) response to current pulse stimulus in the form of transfer function representations for NR excitability. Segments of myelinated internodal (IN) and NR regions are cascaded, representing the remaining downstream axon after a site-of-stimulus introduction of an external current pulse. This cascading network is used to simulate "cable" properties and signal propagation along the length of the axon. This work proposes possible solutions to attenuation losses inherited in the classical myelinated cable models and accounts for neuronal AP velocity as well as introducing signal attenuation and transient delays associated with internodal demyelination. This model could aide as a predictive tool for the diagnosis and analysis of axonal signal integrity associated with demyelination pathology. Possible applications could include functional stimulation control methodologies for axon bundles that may exhibit signal fidelity issues associated with demyelination. It is further proposed that this model may serve as an instructive tool for further development and incorporation of other axon dynamic behaviors such as: relative refractory periods of AP generation, NR AP recovery mechanisms and responses to varied current stimulus input.