Return to search

Raman and near infrared spectroscopic analysis of amniotic fluid : metabolomics of maternal and fetal health indicators

This thesis presents quantitative tools for the metabolomic analysis of amniotic fluid (AF) using vibrational spectroscopy. A total of 300 AF samples were collected for this retrospective cohort study and both Raman and near infrared (NIR) spectra were measured. Spectral data was compressed using a Haar wavelet transform and stage-wise multilinear regression (MLR). Calibration models were calculated for glucose, lactate and uric acid concentrations in AF. Birth weight, gestational diabetes mellitus (GDM) and gestational age were classified with the resulting compressed Raman and NIR spectra, using a genetic algorithm (GA) and a cross-validation approach. Results show that both Raman and NIR spectra of AF were not able to estimate the concentrations of glucose, lactate or uric acid with high precision. However, metabolomic analysis of AF Raman and NIR spectra was capable of estimating the development of GDM, abnormal birth weights as well as gestational ages with sensitivities >75% and specificities >77%. In addition, Raman and NIR metabolomic profiles showed a statistical difference in patients delivering preterm. Of the two spectroscopic analyses studied, NIR spectroscopy of AF has the potential to become a robust and non-invasive diagnostic tool for maternal and fetal health.

Identiferoai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:QMM.112345
Date January 2007
CreatorsPower, Kristin Marie.
PublisherMcGill University
Source SetsLibrary and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada
LanguageEnglish
Detected LanguageEnglish
TypeElectronic Thesis or Dissertation
Formatapplication/pdf
CoverageMaster of Science (Department of Chemistry.)
RightsAll items in eScholarship@McGill are protected by copyright with all rights reserved unless otherwise indicated.
Relationalephsysno: 002712306, proquestno: AAIMR51323, Theses scanned by UMI/ProQuest.

Page generated in 0.0016 seconds