AI can identify regions of tumour unexpectedly predictive of patient survival
Malignant mesothelioma patients are diagnosed and classified by pathologists via tissue biopsy. But only three FDA-approved drugs exist for mesothelioma. The field needs new biomarkers to help identify and classify the disease to improve clinical care and offer possible targets for new drugs.
That’s why Owkin built MesoNet, a machine learning model that predicts the overall survival of malignant mesothelioma patients from digital pathology images. It does this better than the existing subtype classification used by pathologists. In collaboration with expert mesothelioma pathologists, MesoNet identified new biomarkers predictive of prognosis: stroma, tumor cell localization, inflammation, cellular diversity & vascularization. These biomarkers are now under further scientific study to see if we can learn more from them about Mesothelioma.
Explore MesoNet
MesoNet case study: A biomarker of outcome prediction in Mesothelioma