MesoNet

MesoNet is a machine learning model designed to predict overall survival in malignant mesothelioma patients using digital pathology images, aiming to surpass current pathologist-driven subtype classification methods. MesoNet identified new biomarkers related to stroma, tumor cell localization, inflammation, cellular diversity, and vascularization, which are predictive of patient prognosis. This tool has the potential to improve clinical care by aiding prognosis and could lead to the discovery of new drug targets for mesothelioma.