Multi tasks RetinaNet for mitosis detection

TL;DR

This paper enhances mitosis detection in tumor tissues by extending RetinaNet with foreground detection and tumor classification, employing data augmentation to improve robustness across different domains, achieving state-of-the-art results.

Contribution

It introduces a multi-task RetinaNet model with domain generalization techniques for more robust mitosis detection in diverse tumor datasets.

Findings

Achieved state-of-the-art F1 score of 0.5809 on challenging dataset.

Utilized data augmentation to improve domain robustness.

Extended RetinaNet with additional tasks for better performance.

Abstract

The account of mitotic cells is a key feature in tumor diagnosis. However, due to the variability of mitotic cell morphology, it is a highly challenging task to detect mitotic cells in tumor tissues. At the same time, although advanced deep learning method have achieved great success in cell detection, the performance is often unsatisfactory when tested data from another domain (i.e. the different tumor types and different scanners). Therefore, it is necessary to develop algorithms for detecting mitotic cells with robustness in domain shifts scenarios. Our work further proposes a foreground detection and tumor classification task based on the baseline(Retinanet), and utilizes data augmentation to improve the domain generalization performance of our model. We achieve the state-of-the-art performance (F1 score: 0.5809) on the challenging premilary test dataset.