Search

Diagnostic techniques based on optical spectroscopy have the potential to link the biochemical and morphological properties of tissues. Light-induced fluorescence (LIF) spectroscopy as a noninvasive

An aglorithm based on support vector machines (SVM), the most recent advance in pattern recognition, is presented for use in classifying light-induced autofluorescence collected from cancerous and normal tissues. The in vivo autofluorescence spectra used for development and evaluation of SVM diagnostic algorithms were measured from 85 nasopharyngeal carcinoma (NPC) lesions and 131 normal tissue sites from 59 subjects during routine nasal endoscopy. Leave-one-out cross-validation was used to evaluate the performance of the algorithms. An overall diagnostic accuracy of 96\%, a sensitivity of 94\%, and a specificity of 97\% for discriminating nasopharyngeal carcinomas from normal tissues were achieved using a linear SVM algorithm. A diagnostic accuracy of 98\%, a sensitivity of 95\%, and a specificity of 99\% for detecting NPC were achieved with a nonlinear SVM algorithm. In a comparison with previously developed algorithms using the same dataset and the principal component analysis (PCA) technique, the SVM algorithms produced better diagnostic accuracy in all instances. In addition, we investigated a method combining PCA and SVM techniques for reducing the complexity of the SVM algorithms. (C) 2004 Society of Photo-Optical Instrumentation Engineers.

Diagnostic techniques based on optical spectroscopy have the potential to link the biochemical and morphological properties of tissues. Light-induced fluorescence (LIF) spectroscopy as a noninvasive "optical biopsy" method has been used to detect small lesions in vivo using an Hg arc lamp as excitation light. The potential of LIF has been evaluated to improve the accuracy of conventional white light nasal endoscopy. More work has been carried out to improve the existing measuring system with lasers as excitation light. A novel classification method, Support Vector Machines (SVM), has been developed for extracting diagnostic information from autofluorescence spectral signals obtained with nasopharyngeal carcinoma (NPC) and normal tissue. In addition, the possibility to build a simpler algorithm and improve the diagnostic accuracy with the combination of PCA and SVM methods was investigated. It's found that PCA can substantially reduce the complexity of a SVM algorithm without sacrificing the performance of the algorithm. In brief, the classifying performance based on the data in both the spectrum and principal component domains are compatible and excellent; with RBF kernel function, the sensitivity and total predictive accuracy are up to 95.3% and 97.7%, respectively. In the right perspective, the method combining SVM and PCA outperforms other PCA methods. In addition, in order to tracing the autofluorescence spectral signals of tissue layer by layer, a confocal fluorescent spectroscopy system has been set-up. Experiments have been carried out with fluorescent phantom and animal model. With an axial resolution of l0um in tissue, this confocal spectral system observed the spectral differences in spectral shape and spectral peak position among different layers of tissue. In conclusion, light-induced autofluorescence spectroscopy accompanied with robust classification algorithms based on support vector machines provides a noninvasive and feasible tool in the diagnosis o