Ten years of experience with autofluorescence bronchoscopy

Autofluorescence imaging is a modality of imaging of cancerous tissues based on natural phenomenon of emission of light from porphyrins excited by the light and selectively absorbed in neoplasmatic tissues. In our study was performed the analysis of autofluorescence, fluorescence images and fluorescence spectra taken from autofluorescence detection system Xillix Onco-LIFE with spectral camera in changes of upper digestive tract and of changes in bronchial tree. In autofluorescence endoscopy, it is possible to measure autofluorescence intensity ratio (red-to-green), and to perform real-time image processing to improve the quality of imaging and to show places with the highest autofluorescence ratio where the biopsy specimens should be taken. Also data from the red and green channels were extracted, and the correlation between these two components was estimated. All of channels were analysed separately. The second image analysis was performed using CIEL*a*b* colour scale to show the differences between shades of colours in autofluorescence image. CIE L*a*b* is the most complete color space specified by the International Commission on Illumination. It describes the colors visible to the human eye and was created to serve as a device independent model to be used as a reference. In this scale three coordinates represent the lightness of the color, its position between magenta and green and its position between yellow and blue. The aim of the study was an estimation of usefullness of autofluorescence imaging in diagnostics of advance in proliferative changes in upper digestive tract and in bronchial tree. Efectiveness of autofluorescence, fluorescence imaging and analysis of autofluorescence intensity ratio (NCV), fluorescence intensity and intensity of red and green color in gray scale in pictures was performed and all of results were correlated with results from biopsy specimens. Also the analysis of colours intensities, shades of red and green in CIE L*a*b* colour scale and its lumination was performed. All of results were correlated with histopathology of taken biopsy specimens. Results were statistically processed. In this study about 1500 of pictures were taken from changes in bronchial tree. All of pictures were processed with the same mathematical algorithm. In all of changes was observed statistically significantly higher intensity of fluorescence and autofluorescence in changes than in healthy tissue and significantly higher intensity fluorescence and autofluorescence ratio in neoplastic changes than in healthy mucosa. Analysis of colors showed in all of cases significantly lower intensity of green color intensity in neoplasmatic tissues than in healthy tissues. In analysis of red color intensity differences in intensities between neoplasmatic and healthy tissues weren't statistically significant. Also in CIE L*a*b* colour scale analysis allowed to show differencies between healthy and cancerous mucosa even in places not visible in white light and in autofluorescence endoscopy. In bronchial tree autofluorescence imaging allowed to increase specimen biopsy sensitivity WLE vs. AFI vs. AFI + G to 70%, 84%, 94% respectively and WLE + AFI with color analysis to 98%. Autofluorescence imaging allows to estimate the extension of neoplasmatic tissue and allows to estimate healthy and cancerous tissue margins.