Robust Multi-Domain Mitosis Detection

TL;DR

This paper introduces a multi-step mitosis detection algorithm that leverages unpaired image translation to address domain variability in medical imaging, achieving an F1 score of 0.52 on the MIDOG challenge.

Contribution

It proposes a novel use of CycleGAN for domain adaptation in mitosis detection, providing a simple baseline for the MIDOG challenge.

Findings

F1 score of 0.52 on preliminary test set

Effective domain adaptation using CycleGAN

Baseline for MIDOG mitosis detection challenge

Abstract

Domain variability is a common bottle neck in developing generalisable algorithms for various medical applications. Motivated by the observation that the domain variability of the medical images is to some extent compact, we propose to learn a target representative feature space through unpaired image to image translation (CycleGAN). We comprehensively evaluate the performanceand usefulness by utilising the transformation to mitosis detection with candidate proposal and classification. This work presents a simple yet effective multi-step mitotic figure detection algorithm developed as a baseline for the MIDOG challenge. On the preliminary test set, the algorithm scoresan F1 score of 0.52.