MitoDet: Simple and Robust Mitosis Detection

Abstract

Mitotic figure detection is a challenging task in digital pathology that has a direct impact on therapeutic decisions. While automated methods often achieve acceptable results under laboratory conditions, they frequently fail in the clinical deployment phase. This problem can be mainly attributed to a phenomenon called domain shift. An important source of a domain shift is introduced by different microscopes and their camera systems, which noticeably change the colour representation of digitized images. In this method description, we present our submitted algorithm for the Mitosis Domain Generalization Challenge [1], which employs a RetinaNet [5] trained with strong data augmentation and achieves an F1 score of 0.7138 on the preliminary test set.

References

1. 1.Aubreville, M., et al.: Mitosis domain generalization challenge (2021). DOI: https://doi.org/10.5281/zenodo.4573978
2. 2.Cubuk, E.D., Zoph, B., Shlens, J., Le, Q.: RandAugment: practical automated data augmentation with a reduced search space. In: Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M.F., Lin, H. (eds.) Advances in Neural Information Processing Systems, vol. 33, pp. 18613–18624. Curran Associates, Inc. (2020)
3. 3.Ganin Yet al.Domain-adversarial training of neural networksJ. Mach. Learn. Res.2016175913535046191360.68671
4. 4.Li CWang XLiu WLatecki LJDeepMitosis: mitosis detection via deep detection, verification and segmentation networksMed. Image Anal.20184512113310.1016/j.media.2017.12.002
5. 5.Lin, T.Y., Goyal, P., Girshick, R., He, K., Dollar, P.: Focal loss for dense object detection. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 2980–2988 (2017)
6. 6.Marzahl Cet al.Deep learning-based quantification of pulmonary hemosiderophages in cytology slidesSci. Rep.2020101979510.1038/s41598-020-65958-2
7. 7.Müller, S.G., Hutter, F.: TrivialAugment: tuning-free yet state-of-the-art data augmentation. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 774–782 (2021)
8. 8.Smith, L.N., Topin, N.: Super-convergence: very fast training of neural networks using large learning rates. In: Artificial Intelligence and Machine Learning for Multi-Domain Operations Applications, vol. 11006, pp. 369–386. SPIE, May 2019. DOI: https://doi.org/10.1117/12.2520589
9. 9.Sohn, K., Zhang, Z., Li, C.L., Zhang, H., Lee, C.Y., Pfister, T.: A Simple Semi-Supervised Learning Framework for Object Detection. arXiv:2005.04757 [cs], December 2020
10. 10.Stacke KEilertsen GUnger JLundstrom CMeasuring domain shift for deep learning in histopathologyIEEE J. Biomed. Health Inform.202125232533610.1109/JBHI.2020.3032060
11. 11.Tan, M., Le, Q.: EfficientNet: rethinking model scaling for convolutional neural networks. In: Proceedings of the 36th International Conference on Machine Learning, pp. 6105–6114. PMLR, May 2019. ISSN 2640-3498
12. 12.Tellez Det al.Whole-slide mitosis detection in H&E breast histology using PHH3 as a reference to train distilled stain-invariant convolutional networksIEEE Trans. Med. Imaging20183792126213610.1109/TMI.2018.2820199
13. 13.Xie, Q., Luong, M.T., Hovy, E., Le, Q.V.: Self-training with noisy student improves ImageNet classification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 10687–10698 (2020)
14. 14.Zlocha MDou QGlocker Bet al.Shen Det al.Improving RetinaNet for CT lesion detection with dense masks from weak RECIST labelsMedical Image Computing and Computer Assisted Intervention – MICCAI 20192019ChamSpringer40241010.1007/978-3-030-32226-7_45

Index Terms

• Applied computing

  • Life and medical sciences

    • Computational biology

    • Genetics

• Computing methodologies

  • Artificial intelligence

    • Computer vision

      • Computer vision problems

  • Machine learning