The growing demand for liver transplantation is leading medicine to explore new technologies in order to extend the field of viable transplants. Automatic hepatic steatosis assessment is the first step towards the development of a computeraided liver diagnosis due to its importance as a risk factor for primary dysfunction. Color and texture are considered as two fundamental visual characteristics to assess the degree of hepatic steatosis. The aim of this study is to determine the discriminating features for liver image classification according to three steatosis classes: mild, moderate or severe. First, color features are extracted in three color spaces namely: RGB, HSV and YCbCr. Texture features are then extracted from RGB components using co-occurrence matrices, Local Binary Pattern (LBP) and Local Phase Quantization (LPQ). Additionally, feature-extraction was followed by generating a linear model regression using Least Absolute Shrinkage and Selection Operator (LASSO) allowing the classification task. The experimental results show that RGB histograms provide better classification accuracy. However, the relevance of these preliminary results is not sufficient to draw a decisive conclusion due to the experimental database, which is really small, very unbalanced and the images were captured under unknown and widely varying conditions.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.