Options
Excitation-emission spectroscopy studies of human colorectal tissues for improving cancer diagnosis using tissue autofluorescence
Author
Zhao, Tianyun
Supervisor
Chia, Teck Chee
Abstract
This work aims to optimize the excitation wavelength for fluorescence spectroscopy to get better human colorectal cancer diagnosis. The autofluorescence spectra of colorectal tissue are studied over a wide excitation range (220 nm, and 350-600 nm) using a laboratory spectrofluorophotometer (RF-5301PC instrument, Shimadzu Corporation, Kyoto, Japan). 22 samples from different patient are used in this investigation.
Excitation wavelength at UV light is powerful light to study the autofluorescence spectra and to identify the fluorophores inside human tissues. Using excitation wavelength 220 nm enables us to know the main autofluorescence characteristics of the tissue and the main fluorescence peaks located within the tissue spectra. And according to the characteristics of tissues spectra under all excitation wavelengths, we choose two effective excitation wavelengths for further research.
Excitation wavelengths : 350 nm and 470 nm are identified as effective excitation wavelengths for study of human colorectal tissues' fluorescence. Both of them induce the best one-prominent-peak shape of normal human colorectal tissue autofluorescence and the two-peak shape of cancerous human colorectal tissue autofluorescence. The two peaks of cancerous spectra are located at 510 nm and 610 nm and have similar intensity. The use of excitation wavelength 470 nm is a breakthrough in the using of visible light in photodiagnosis, and it also provides high sensitivity and specificity in the discrimination of normal and cancerous tissues. Instead of using intensity alone, the ratios of the peak intensity of emission fluorescence at different selective wavelengths were used to discriminate between normal and cancerous tissues. The use of peak intensity ratio avoids the influences such as the incident angle of light, the change of light source intensity, light source distance from sample surface and distance to emission monochromator, and systematic errors of the spectrofluorometer in the measurements. Several analytic methods such as ratio of peak intensity (I510/I610...), ratio of peak area (A510/A610...) and spectra area in green region to the whole spectra area (Ag/Aw) under excitation 350 nm and 470 nm are compared and have provided high sensitivity and specificity in the discrimination of normal and cancerous human colorectal tissues.
Peak deconvolution is a method to adjust parameters of a set of curves with input peaks and sort at the super position of the curves to fit peaks in a real spectrum. Deconvolution of typical normal and cancerous human colorectal tissue autofluorescence spectra provides us more information about the autofluorescence that we measured. The observed fluorescence is the combination of fluorescence from surface and inner layer of the tissues and the tissue fluorescence spectra consist of linear contributions of fluorescence from individual tissue fluorophores. We believe that each peakD (decomposed peak) corresponds to each fluorophore in tissues, such as peakD 470 reflects fluorophore of NADH, which exists in both normal and cancerous tissues but its contribution to the whole spectrum has changed. PeakD 610 reflects porphyrin in cancerous human tissues. Their contributions are different in normal and cancerous tissues. We believe that peak deconvolution of the observed fluorescence spectrum is a mean to understand the origins of tissue fluorescence and can provide critical information by detect biochemical changes in the malignant process.
Excitation wavelength at UV light is powerful light to study the autofluorescence spectra and to identify the fluorophores inside human tissues. Using excitation wavelength 220 nm enables us to know the main autofluorescence characteristics of the tissue and the main fluorescence peaks located within the tissue spectra. And according to the characteristics of tissues spectra under all excitation wavelengths, we choose two effective excitation wavelengths for further research.
Excitation wavelengths : 350 nm and 470 nm are identified as effective excitation wavelengths for study of human colorectal tissues' fluorescence. Both of them induce the best one-prominent-peak shape of normal human colorectal tissue autofluorescence and the two-peak shape of cancerous human colorectal tissue autofluorescence. The two peaks of cancerous spectra are located at 510 nm and 610 nm and have similar intensity. The use of excitation wavelength 470 nm is a breakthrough in the using of visible light in photodiagnosis, and it also provides high sensitivity and specificity in the discrimination of normal and cancerous tissues. Instead of using intensity alone, the ratios of the peak intensity of emission fluorescence at different selective wavelengths were used to discriminate between normal and cancerous tissues. The use of peak intensity ratio avoids the influences such as the incident angle of light, the change of light source intensity, light source distance from sample surface and distance to emission monochromator, and systematic errors of the spectrofluorometer in the measurements. Several analytic methods such as ratio of peak intensity (I510/I610...), ratio of peak area (A510/A610...) and spectra area in green region to the whole spectra area (Ag/Aw) under excitation 350 nm and 470 nm are compared and have provided high sensitivity and specificity in the discrimination of normal and cancerous human colorectal tissues.
Peak deconvolution is a method to adjust parameters of a set of curves with input peaks and sort at the super position of the curves to fit peaks in a real spectrum. Deconvolution of typical normal and cancerous human colorectal tissue autofluorescence spectra provides us more information about the autofluorescence that we measured. The observed fluorescence is the combination of fluorescence from surface and inner layer of the tissues and the tissue fluorescence spectra consist of linear contributions of fluorescence from individual tissue fluorophores. We believe that each peakD (decomposed peak) corresponds to each fluorophore in tissues, such as peakD 470 reflects fluorophore of NADH, which exists in both normal and cancerous tissues but its contribution to the whole spectrum has changed. PeakD 610 reflects porphyrin in cancerous human tissues. Their contributions are different in normal and cancerous tissues. We believe that peak deconvolution of the observed fluorescence spectrum is a mean to understand the origins of tissue fluorescence and can provide critical information by detect biochemical changes in the malignant process.
Date Issued
2003
Call Number
RC268.4 Zha
Date Submitted
2003