Semester

Spring

Date of Graduation

2002

Document Type

Dissertation

Degree Type

PhD

College

Statler College of Engineering and Mineral Resources

Department

Chemical and Biomedical Engineering

Committee Chair

Stephanie Caswell Schuckers.

Abstract

Sudden infant death syndrome (SIDS) is the sudden death of an infant under 1 year old without any apparent sign ahead of time. It is the leading cause of death of infants between the age of one month and one year in the developed countries. After thirty years substantial worldwide research, it is still impossible to predict or prevent SIDS. With the long-term goal to prevent SIDS, this study explores prediction of life-threatening events using heart rate variability (HRV). HRV is an indirect measure of the autonomic nervous system input to the cardiovascular system. The hypothesis is that infants who are at high risk for life-threatening events and SIDS have a lack of nervous system control. Although statistically significant differences in HRV have been found between infants at high risk for SIDS and controls in previous studies, few studies go one step further to classify the infant group or put effort toward prediction of future events for a specific infant.;We investigate whether HRV can differentiate infants at risk for future apparent life-threatening events (FALTE) using the dataset collected by the Collaborative Home Infant Monitoring Evaluation (CHIME) study group. To achieve this, we developed and validated an artifact rejection routine, used statistical methods to study differences between normal and FALTE infant groups, and explored the relationship between HRV parameters. Lastly, the major work of our study is to design a model to classify infants using HRV, including logistic regression, decision tree, and neural networks. Promising results (67% sensitivity and 100% specificity) are achieved for the test dataset by varying the model parameters and the threshold. Up until now, there is little information available for determining the likelihood of future life-threatening events for a newborn. If prediction at this accuracy level can be achieved through the measurement of HRV, infants at risk for life-threatening events can be identified and assisted by offering corresponding therapy.

Download

Share

COinS