Publication Date
2010
Document Type
Thesis
Committee Members
Ping He (Committee Member), Michael Raymer (Committee Member), Nicholas Reo (Committee Co-chair), Julie Skipper (Committee Co-chair)
Degree Name
Master of Science in Engineering (MSEgr)
Abstract
Baseline correction of NMR-based metabolomic spectra is a key step in data processing to elucidate biomarkers of diseased and toxic states. Automated baseline correction methods often use human-selected parameters. Presented is a method for automated baseline correction using parameters selected from and without manipulation of the spectrum. The focus is on generating an accurate signal map based on the differentiating characteristics of baseline noise and signal. The presented method of signal map generation and baseline correction was developed and tested on 13C and 1H NMR spectra. The spectra were as simple as chemical standards containing less than 25 signal peaks to very dense urinary metabolite profiles. This method has shown proper correction of even the most complicated NMR biofluid spectra and is acceptable for use in multivariate analyses. This technique may also be robust enough to utilize in other spectroscopic methods in which thermally generated baseline noise is present.
Page Count
69
Department or Program
Department of Biomedical, Industrial & Human Factors Engineering
Year Degree Awarded
2010
Copyright
Copyright 2010, all rights reserved. This open access ETD is published by Wright State University and OhioLINK.
