This pipeline segments cell nuclei into one of 5 different classes: Neoplastic, Inflammatory, Connective, Dead, Non-Neoplastic Epithelial.
The model is trained on the PanNuke dataset using the HoVerNet architecture.
The model is part of the publication
"A Pragmatic Machine Learning Approach to Quantify Tumor-Infiltrating Lymphocytes in Whole Slide Images" Cancers 2022 https://doi.org/10.3390/cancers14122974
Authors: Nikita Shvetsov, Morten Grønnesby, Edvard Pedersen, Kajsa Møllersen, Lill-Tove Rasmussen Busund, Ruth Schwienbacher, Lars Ailo Bongo, Thomas Karsten Kilvaer
Copyrights: UiT - The Arctic University of Norway
CC SA-BY-NC 4.0