Skip to main content Skip to navigation

CoNIC Challenge

Page Contents

  1. About
  2. Method manuscripts
  3. Participant docker containers
  4. WSI-level results


Nuclear segmentation, classification and quantification within Haematoxylin & Eosin stained histology images enables the extraction of interpretable cell-based features that can be used in downstream explainable models in computational pathology (CPath). To help drive forward research and innovation for automatic nuclei recognition in CPath, we organise the Colon Nuclei Identification and Counting (CoNIC) Challenge. The challenge requires researchers to develop algorithms that perform segmentation, classification and counting of 6 different types of nuclei within the current largest known publicly available nuclei-level dataset in CPath, containing around half a million labelled nuclei.

The challenge consisted of two tasks:

  1. Nuclear segmentation and classification
    • Automatically localise the boundaries of different types of nuclei
  2. Cellular composition
    • Automatically predict the counts of different types of nuclei

Visit the challenge website, where you can find instructions on downloading the main dataset.

Method Manuscripts

Below we provide method description manuscripts that described each of the submissions to the challenge.

Team Name Task One Position Task Two Position
EPFL | StarDist 1st 3rd
Pathology AI 3rd 1st
AI_medical 5th 2nd
MDC Berlin | IFP Bern 2nd 9th
CIA Group 7th 4th
LSL000UD 4th 6th
Softsensor_Group 12th 5th
Arontier 6th 7th
MBZUAI_CoNiC 10th 8th
MAIIA 8th 18th
ciscNet 9th 11th
Denominator 11th 10th
BMS_LAB 13th 15th
GDPH_HC 14th 13th
TIA Warwick 24th 14th
conic-challenge-inescteam 15th N/A
Jiffy Labs and CET CV Lab 23rd 16th
Aman 16th N/A
DH-Goods 19th 17th
Bin 17th 20th
Sk 21st 19th
VNIT 20th 22nd
QuIL N/A 23rd
IVG 26th 24th

    Participant Docker Containers

    To download Docker containers from the top challenge algorithms, click here.

    WSI-Level Results

    To download WSI-level results obtained after processing ~1.7K WSIs with the winning algorithms on each task of the challenge click here.