Improving Mitosis Detection via UNet-Based Adversarial Domain Homogenizer

Tirupati Chandra, Sahar Nasser, Nikhil Kurian, Amit Sethi

2023

Abstract

The effective counting of mitotic figures in cancer pathology specimen is a critical task for deciding tumor grade and prognosis. Automated mitosis detection through deep learning-based image analysis often fails on unseen patient data due to domain shifts in the form of changes in stain appearance, pixel noise, tissue quality, and magnification. This paper proposes a domain homogenizer for mitosis detection that attempts to alleviate domain differences in histology images via adversarial reconstruction of input images. The proposed homogenizer is based on a U-Net architecture and can effectively reduce domain differences commonly seen with histology imaging data. We demonstrate our domain homogenizer’s effectiveness by showing a reduction in domain differences between the preprocessed images. Using this homogenizer with a RetinaNet object detector, we were able to outperform the baselines of the 2021 MIDOG challenge in terms of average precision of the detected mitotic figures.

Paper Citation

in Harvard Style

Chandra T., Nasser S., Kurian N. and Sethi A. (2023). Improving Mitosis Detection via UNet-Based Adversarial Domain Homogenizer. In Proceedings of the 16th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2023) - Volume 2: BIOIMAGING; ISBN 978-989-758-631-6, SciTePress, pages 52-56. DOI: 10.5220/0011629700003414

in Bibtex Style

@conference{bioimaging23,
author={Tirupati Chandra and Sahar Nasser and Nikhil Kurian and Amit Sethi},
title={Improving Mitosis Detection via UNet-Based Adversarial Domain Homogenizer},
booktitle={Proceedings of the 16th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2023) - Volume 2: BIOIMAGING},
year={2023},
pages={52-56},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0011629700003414},
isbn={978-989-758-631-6},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 16th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2023) - Volume 2: BIOIMAGING
TI - Improving Mitosis Detection via UNet-Based Adversarial Domain Homogenizer
SN - 978-989-758-631-6
AU - Chandra T.
AU - Nasser S.
AU - Kurian N.
AU - Sethi A.
PY - 2023
SP - 52
EP - 56
DO - 10.5220/0011629700003414
PB - SciTePress