CoNIC Challenge: Pushing the Frontiers of Nuclear Detection, Segmentation, Classification and Counting
Abstract
Nuclear detection, segmentation and morphometric profiling are essential in helping us further understand the relationship between histology and patient outcome. To drive innovation in this area, we setup a community-wide challenge using the largest available dataset of its kind to assess nuclear segmentation and cellular composition. Our challenge, named CoNIC, stimulated the development of reproducible algorithms for cellular recognition with real-time result inspection on public leaderboards. We conducted an extensive post-challenge analysis based on the top-performing models using 1,658 whole-slide images of colon tissue. With around 700 million detected nuclei per model, associated features were used for dysplasia grading and survival analysis, where we demonstrated that the challenge's improvement over the previous state-of-the-art led to significant boosts in downstream performance. Our findings also suggest that eosinophils and neutrophils play an important role in the tumour microevironment. We release challenge models and WSI-level results to foster the development of further methods for biomarker discovery.
Cite
@article{arxiv.2303.06274,
title = {CoNIC Challenge: Pushing the Frontiers of Nuclear Detection, Segmentation, Classification and Counting},
author = {Simon Graham and Quoc Dang Vu and Mostafa Jahanifar and Martin Weigert and Uwe Schmidt and Wenhua Zhang and Jun Zhang and Sen Yang and Jinxi Xiang and Xiyue Wang and Josef Lorenz Rumberger and Elias Baumann and Peter Hirsch and Lihao Liu and Chenyang Hong and Angelica I. Aviles-Rivero and Ayushi Jain and Heeyoung Ahn and Yiyu Hong and Hussam Azzuni and Min Xu and Mohammad Yaqub and Marie-Claire Blache and Benoît Piégu and Bertrand Vernay and Tim Scherr and Moritz Böhland and Katharina Löffler and Jiachen Li and Weiqin Ying and Chixin Wang and Dagmar Kainmueller and Carola-Bibiane Schönlieb and Shuolin Liu and Dhairya Talsania and Yughender Meda and Prakash Mishra and Muhammad Ridzuan and Oliver Neumann and Marcel P. Schilling and Markus Reischl and Ralf Mikut and Banban Huang and Hsiang-Chin Chien and Ching-Ping Wang and Chia-Yen Lee and Hong-Kun Lin and Zaiyi Liu and Xipeng Pan and Chu Han and Jijun Cheng and Muhammad Dawood and Srijay Deshpande and Raja Muhammad Saad Bashir and Adam Shephard and Pedro Costa and João D. Nunes and Aurélio Campilho and Jaime S. Cardoso and Hrishikesh P S and Densen Puthussery and Devika R G and Jiji C and Ye Zhang and Zijie Fang and Zhifan Lin and Yongbing Zhang and Chunhui Lin and Liukun Zhang and Lijian Mao and Min Wu and Vi Thi-Tuong Vo and Soo-Hyung Kim and Taebum Lee and Satoshi Kondo and Satoshi Kasai and Pranay Dumbhare and Vedant Phuse and Yash Dubey and Ankush Jamthikar and Trinh Thi Le Vuong and Jin Tae Kwak and Dorsa Ziaei and Hyun Jung and Tianyi Miao and David Snead and Shan E Ahmed Raza and Fayyaz Minhas and Nasir M. Rajpoot},
journal= {arXiv preprint arXiv:2303.06274},
year = {2023}
}