Robust Multi-Domain Mitosis Detection
This work addresses domain generalization for medical imaging, specifically in mitosis detection, but is incremental as it builds on existing methods for a baseline challenge.
The paper tackled domain variability in medical image analysis by using CycleGAN for unpaired image translation to improve mitosis detection, achieving an F1 score of 0.52 on a preliminary test set.
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.