Deep Learning–Based Image Analysis of Liver Steatosis in Mouse Models

Abstract

The incidence of non-alcoholic fatty liver disease (NAFLD) is a continuously growing health problem worldwide, along with obesity. Therefore, both novel methods to efficiently study the manifestation of NAFLD and to analyze drug efficacy in pre-clinical models are needed. In the present study, we developed a deep neural network -based model to quantify micro- and macrovesicular steatosis in the liver on hematoxylin-eosin stained whole slide images (WSIs), using the cloud-based platform, Aiforia Create (Aiforia Technologies, Helsinki, Finland). The training data included a total of 101 WSIs from dietary interventions of wild-type mice and from two genetically modified (GM) mouse models with steatosis. The algorithm was trained for the following: to detect liver parenchyma, to exclude the blood vessels and any artefacts generated during tissue processing and image acquisition, to recognize and differentiate the areas of micro- and macrovesicular steatosis, and to quantify the recognized tissue area. The results of the image analysis replicated well the evaluation by expert pathologists, and correlated well with the liver fat content measured by EcoMRI ex vivo, and the correlation with total liver triglycerides were notable. In conclusion, the developed deep learning-based model is a novel tool for studying liver steatosis in mouse models on paraffin sections, and thus, can facilitate reliable quantification of the amount of steatosis in large preclinical study cohorts.