Return to search

Statistical Classification of Brainstem Auditory Evoked Potentials and The Effects of Local Cooling on Canine Spinal Cord Blood Flow

<p>Brainstem auditory evoked potentials (BAEPs), the time locked scalp recorded electrical responses to auditory stimuli are clinically useful for diagnosing and monitoring the disorders of neurological system. The first part of the thesis presents the development and evaluation of computer based statistical classifiers to recognize normal and patients' BAEPs. The classifiers using time domain features; namely latencies of peaks V, IV and III of BAEPs, gave the optimal performance as measured by accuracy (85.3%), sensitivity (80%) among other indices. Power spectra of BAEPs in normals show three main frequency bands. In a second formulation of classifiers, three features, each feature representing the total power in each of the three frequency bands were used. The classifiers designed with these frequency domain features yielded a performance with accuracies upto 77%. Both of these classifiers may be used to assist the clinician while assessing the BAEPs.</p> <p>The second part of the thesis examines the effects of cooling on canine spinal cord blood flow (SCBF). The SCBF is an important physiological variable altered during spinal cord injury (SCI). Localized cooling of the spinal cord improves functional recovery after SCI. Hence, the effects of cooling on SCBF in normal canine cord were studied. A computerized system to measure the SCBF using hydrogen polarography at two control sites and two cooled sites was developed. In five dogs, SCBF decreased to 50% of the normothermic values during cooling of the cord to a temperature of 16 degrees Celsius. The SCBF did not change at control sites. It is postulated that following SCI decreased SCBF due to localized cooling tends to prevent the outpouring of edema fluid and other toxic factors from the injured vessels, preserving the cord function. Thus, the finding of decreased SCBF during spinal cord cooling is clinically relevant.</p> / Doctor of Philosophy (PhD)

Identiferoai:union.ndltd.org:mcmaster.ca/oai:macsphere.mcmaster.ca:11375/6855
Date07 1900
CreatorsKamath, Venkatraya Markad
ContributorsGhista, Dhanjoo N., Neuroscience
Source SetsMcMaster University
Detected LanguageEnglish
Typethesis

Page generated in 0.0021 seconds