Development of a segmentation method for crypts of Lieberkühn detection using autofluorescence colon images
Abstract
Colorectal cancer is one of the diseases with the greatest impact worldwide. It is very well known that the pathology prognosis is linked to an early detection and subsequent treatment. The processing of colon autofluorescence images allows the detection and segmentation of the pits from the crypts of Lieberkühn. The procedure is done by using information coming from endogenous fluorophores. The segmentation is essential to automatically quantify the type and distribution of the pits in each tissue according the patterns established by Kudo. Based on that pattern, the pathology is staged. Therefore, a method was designed to segment the pits from colon tissue images obtained by widefield autofluorescence microscopy.