Oliver, KV; (2015) Infrared spectroscopy as a clinical diagnostic method for detection of disease states: developments and applications in kidney diseases and cancer diagnoses. Doctoral thesis , UCL (University College London).

Preview

Text K_Oliver_Thesis.pdf Download (39MB) | Preview

Abstract

ATR-FTIR spectroscopy has been used to develop rapid and accurate methods for urine analyses and cancer diagnoses. Complications associated with analyses of dried samples by FTIR spectroscopy were found to arise because of the existence of partial hydration states of some substances. Recognition and accommodation of such states is necessary to ensure precise deconvolution of spectra. Dramatic hydration effects were observed in the IR spectrum of urea. Four forms of urea were identified in IR spectra, corresponding to solution (1468 cm-1) and dry (1464 cm-1) states and two intermediate (1454 cm-1 and 1443 cm-1) forms that arise from specific urea-water and/or urea-urea interactions. Less pronounced hydration effects were also identified in the spectra of creatinine and other urinary compounds. ATR-FTIR spectroscopy was then applied to diagnosis of the kidney stone disease, cystinuria. A simple method for detecting and quantitating urinary insoluble cystine was developed based on the integrated area of the cystine 1296 cm-1 band. Clinically relevant levels could be detected in ATR-FTIR spectra of dried insoluble fractions of urine samples without any requirement for chemical manipulation. The use of ATR-FTIR spectroscopy as a point-of-care diagnostic tool for kidney diseases was investigated by installation of a portable benchtop spectrometer in a hospital clinic. Nurses recorded spectra of 295 unprocessed urine samples and the data were correlated with clinical diagnoses. Initial results suggested that differences in the integrated areas of key spectral features could discriminate between healthy patients and those with chronic kidney disease (CKD), and indicate the extent of kidney damage. ATR-FTIR spectra of 80 pancreatic tissue samples were also recorded and analysed for spectral patterns associated with cancer using PLS logistic regression. Tumour tissue spectra could be separated from normal tissue spectra with 90 % sensitivity. Tumour samples could be further separated into cancer stages with accuracies of 64-84 %.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item