OBJECTIVES: The aim of this work was to verify the differentiation between normal and pathological human carotid artery tissues by using fluorescence and reflectance spectroscopy in the 400- to 700-nm range and the spectral characterization by means of principal components analysis. BACKGROUND DATA: Atherosclerosis is the most common and serious pathology of the cardiovascular system. Principal components represent the main spectral characteristics that occur within the spectral data and could be used for tissue classification. MATERIALS AND METHODS: Sixty postmortem carotid artery fragments (26 non-atherosclerotic and 34 atherosclerotic with non-calcified plaques) were studied. The excitation radiation consisted of a 488-nm argon laser. Two 600-microm core optical fibers were used, one for excitation and one to collect the fluorescence radiation from the samples. The reflectance system was composed of a halogen lamp coupled to an excitation fiber positioned in one of the ports of an integrating sphere that delivered 5 mW to the sample. The photo-reflectance signal was coupled to a (1/4)-m spectrograph via an optical fiber. Euclidean distance was then used to classify each principal component score into one of two classes, normal and atherosclerotic tissue, for both fluorescence and reflectance. RESULTS: The principal components analysis allowed classification of the samples with 81% sensitivity and 88% specificity for fluorescence, and 81% sensitivity and 91% specificity for reflectance. CONCLUSIONS: Our results showed that principal components analysis could be applied to differentiate between normal and atherosclerotic tissue with high sensitivity and specificity.
OBJECTIVES: The aim of this work was to verify the differentiation between normal and pathological human carotid artery tissues by using fluorescence and reflectance spectroscopy in the 400- to 700-nm range and the spectral characterization by means of principal components analysis. BACKGROUND DATA: Atherosclerosis is the most common and serious pathology of the cardiovascular system. Principal components represent the main spectral characteristics that occur within the spectral data and could be used for tissue classification. MATERIALS AND METHODS: Sixty postmortem carotid artery fragments (26 non-atherosclerotic and 34 atherosclerotic with non-calcified plaques) were studied. The excitation radiation consisted of a 488-nm argon laser. Two 600-microm core optical fibers were used, one for excitation and one to collect the fluorescence radiation from the samples. The reflectance system was composed of a halogen lamp coupled to an excitation fiber positioned in one of the ports of an integrating sphere that delivered 5 mW to the sample. The photo-reflectance signal was coupled to a (1/4)-m spectrograph via an optical fiber. Euclidean distance was then used to classify each principal component score into one of two classes, normal and atherosclerotic tissue, for both fluorescence and reflectance. RESULTS: The principal components analysis allowed classification of the samples with 81% sensitivity and 88% specificity for fluorescence, and 81% sensitivity and 91% specificity for reflectance. CONCLUSIONS: Our results showed that principal components analysis could be applied to differentiate between normal and atherosclerotic tissue with high sensitivity and specificity.