How Distance Transform Maps Boost Segmentation CNNs: An Empirical Study

Jun Ma, Zhan Wei, Yiwen Zhang, Yixin Wang, Rongfei Lv, Cheng Zhu, Chen Gaoxiang, Jianan Liu, Chao Peng, Lei Wang, Yunpeng Wang, Jianan Chen
Proceedings of the Third Conference on Medical Imaging with Deep Learning, PMLR 121:479-492, 2020.

Abstract

Incorporating distance transform maps of ground truth into segmentation CNNs has been an interesting new trend in the last year. Despite many great works leading to improvements on a variety of segmentation tasks, the comparison among these methods has not been well studied. In this paper, our {\em first contribution} is to summarize the latest developments of these methods in the 3D medical segmentation field. The {\em second contribution} is that we systematically evaluated five benchmark methods on two representative public datasets. These experiments highlight that all the five benchmark methods can bring performance gains to baseline V-Net. However, the implementation details have a noticeable impact on the performance, and not all the methods hold the benefits on different datasets. Finally, we suggest the best practices and indicate unsolved problems for incorporating distance transform maps into CNNs, which we hope would be useful for the community. The codes and trained models are publicly available at: {https://github.com/JunMa11/SegWithDistMap}.

Cite this Paper

BibTeX
@InProceedings{pmlr-v121-ma20b, title = {How Distance Transform Maps Boost Segmentation CNNs: An Empirical Study}, author = {Ma, Jun and Wei, Zhan and Zhang, Yiwen and Wang, Yixin and Lv, Rongfei and Zhu, Cheng and Gaoxiang, Chen and Liu, Jianan and Peng, Chao and Wang, Lei and Wang, Yunpeng and Chen, Jianan}, booktitle = {Proceedings of the Third Conference on Medical Imaging with Deep Learning}, pages = {479--492}, year = {2020}, editor = {Arbel, Tal and Ben Ayed, Ismail and de Bruijne, Marleen and Descoteaux, Maxime and Lombaert, Herve and Pal, Christopher}, volume = {121}, series = {Proceedings of Machine Learning Research}, month = {06--08 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v121/ma20b/ma20b.pdf}, url = {https://proceedings.mlr.press/v121/ma20b.html}, abstract = {Incorporating distance transform maps of ground truth into segmentation CNNs has been an interesting new trend in the last year. Despite many great works leading to improvements on a variety of segmentation tasks, the comparison among these methods has not been well studied. In this paper, our {\em first contribution} is to summarize the latest developments of these methods in the 3D medical segmentation field. The {\em second contribution} is that we systematically evaluated five benchmark methods on two representative public datasets. These experiments highlight that all the five benchmark methods can bring performance gains to baseline V-Net. However, the implementation details have a noticeable impact on the performance, and not all the methods hold the benefits on different datasets. Finally, we suggest the best practices and indicate unsolved problems for incorporating distance transform maps into CNNs, which we hope would be useful for the community. The codes and trained models are publicly available at: {https://github.com/JunMa11/SegWithDistMap}.} }
Endnote
%0 Conference Paper %T How Distance Transform Maps Boost Segmentation CNNs: An Empirical Study %A Jun Ma %A Zhan Wei %A Yiwen Zhang %A Yixin Wang %A Rongfei Lv %A Cheng Zhu %A Chen Gaoxiang %A Jianan Liu %A Chao Peng %A Lei Wang %A Yunpeng Wang %A Jianan Chen %B Proceedings of the Third Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2020 %E Tal Arbel %E Ismail Ben Ayed %E Marleen de Bruijne %E Maxime Descoteaux %E Herve Lombaert %E Christopher Pal %F pmlr-v121-ma20b %I PMLR %P 479--492 %U https://proceedings.mlr.press/v121/ma20b.html %V 121 %X Incorporating distance transform maps of ground truth into segmentation CNNs has been an interesting new trend in the last year. Despite many great works leading to improvements on a variety of segmentation tasks, the comparison among these methods has not been well studied. In this paper, our {\em first contribution} is to summarize the latest developments of these methods in the 3D medical segmentation field. The {\em second contribution} is that we systematically evaluated five benchmark methods on two representative public datasets. These experiments highlight that all the five benchmark methods can bring performance gains to baseline V-Net. However, the implementation details have a noticeable impact on the performance, and not all the methods hold the benefits on different datasets. Finally, we suggest the best practices and indicate unsolved problems for incorporating distance transform maps into CNNs, which we hope would be useful for the community. The codes and trained models are publicly available at: {https://github.com/JunMa11/SegWithDistMap}.
APA
Ma, J., Wei, Z., Zhang, Y., Wang, Y., Lv, R., Zhu, C., Gaoxiang, C., Liu, J., Peng, C., Wang, L., Wang, Y. & Chen, J.. (2020). How Distance Transform Maps Boost Segmentation CNNs: An Empirical Study. Proceedings of the Third Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 121:479-492 Available from https://proceedings.mlr.press/v121/ma20b.html.

Related Material