Machine learning-based collagen fiber quantification enables analysis of the pancreatic cancer tumor microenvironment directly from hematoxylin and eosin-stained whole slide images

Study Background

While immunotherapy has provided benefit to many cancer patients, some cancer subtypes, such as pancreatic ductal adenocarcinoma (PDAC), are notoriously resistant to these approaches [1].

The desmoplastic stroma of PDAC may act as a barrier against T-cell infiltration [2]. However, the relationship between collagen and anti-tumor immunity is complex, as stromal collagen has been proposed to restrain PDAC progression [3].

To investigate the implications of collagen in PDAC and other cancer types, we developed a machine learning (ML) method to extract and quantify collagen features directly from hematoxylin and eosin (H&E)-stained whole slide images (WSI).

Conference

SITC 2024