Segmentation of Pathological Features of Rat Bile Duct Carcinoma from Hyperspectral Images

Abstract

Liver disease has always been one of the key concerns of the medical community. In recent years, liver diagnostic techniques such as serology have been continuously developed, but the pathological diagnosis of the liver, especially the advancement of liver biopsy technology, is still the most reliable basis for the diagnosis and treatment of liver diseases. In this paper, pathological sections of rat bile duct carcinoma were used as experimental samples to identify and analyze liver tumor by microscopic hyperspectral imaging (MHSI)technique. The proposed method combines the Otsu (OTSU)algorithm with the support vector machine (SVM)algorithm to segment the liver tumor, and comparing with SVM segmentation results. Experimental results show that the OTSU-SVM method has an accuracy of 94.59%, which provides a potential reference value for the pathological diagnosis of liver tumors.