# SegLoss

**Repository Path**: allces/SegLoss

## Basic Information

- **Project Name**: SegLoss
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2021-09-13
- **Last Updated**: 2021-09-13

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Loss functions for image segmentation
![A collection of loss functions for medical image segmentation](https://github.com/JunMa11/SegLoss/blob/master/test/LossOverview.PNG)


```
@article{LossOdyssey,
title = {Loss Odyssey in Medical Image Segmentation},
journal = {Medical Image Analysis},
volume = {71},
pages = {102035},
year = {2021},
author = {Jun Ma and Jianan Chen and Matthew Ng and Rui Huang and Yu Li and Chen Li and Xiaoping Yang and Anne L. Martel}
doi = {https://doi.org/10.1016/j.media.2021.102035},
url = {https://www.sciencedirect.com/science/article/pii/S1361841521000815}
}
```

**Take-home message: compound loss functions are the most robust losses, especially for the highly imbalanced segmentation tasks.**

> Some recent side evidence: [the winner](https://link.springer.com/chapter/10.1007/978-3-030-67194-5_4) in MICCAI 2020 [HECKTOR](http://www.aicrowd.com/challenges/hecktor) Challenge used DiceFocal loss; the [winner and runner-up](https://arxiv.org/pdf/2101.00232) in MICCAI 2020 [ADAM](http://adam.isi.uu.nl/) Challenge used DiceTopK loss.


|Date|First Author|Title|Conference/Journal|
|---|---|---|---|
|20210418|Bingyuan Liu|The hidden label-marginal biases of segmentation losses [(pytorch)](https://github.com/by-liu/SegLossBias) | [arxiv](https://arxiv.org/abs/2104.08717)|
|20210330|Suprosanna Shit and Johannes C. Paetzold|**clDice** - a Novel Topology-Preserving Loss Function for Tubular Structure Segmentation [(keras and pytorch)](https://github.com/jocpae/clDice)|[CVPR 2021](https://arxiv.org/abs/2003.07311)|
|20210325|Attila Szabo, Hadi Jamali-Rad|Tilted Cross Entropy (TCE): Promoting Fairness in Semantic Segmentation|[arxiv](https://arxiv.org/abs/2103.14051v1)|
|20210318|Xiaoling Hu|Topology-Aware Segmentation Using Discrete Morse Theory [arxiv](https://arxiv.org/abs/2103.09992v1)|[ICLR 2021](https://openreview.net/forum?id=LGgdb4TS4Z)|
|20210211|Hoel Kervadec|Beyond pixel-wise supervision: semantic segmentation with higher-order shape descriptors|[Submitted to MIDL 2021](https://openreview.net/forum?id=nqe6e0oJ_fL)|
|20210210|Rosana EL Jurdi|A Surprisingly Effective **Perimeter-based Loss** for Medical Image Segmentation|[Submitted to MIDL 2021](https://openreview.net/forum?id=NDEmtyb4cXu)|
|20201222|Zeju Li|Analyzing Overfitting Under Class Imbalance in Neural Networks for Image Segmentation|[TMI](https://ieeexplore.ieee.org/document/9302891)|
|20210129|Nick Byrne|A Persistent Homology-Based **Topological Loss** Function for Multi-class CNN Segmentation of Cardiac MRI [arxiv](https://arxiv.org/abs/2008.09585)| [STACOM 2020](https://link.springer.com/chapter/10.1007/978-3-030-68107-4_1)|
|20201019|Hyunseok Seo|Closing the Gap Between Deep Neural Network Modeling and Biomedical Decision-Making Metrics in Segmentation via **Adaptive Loss** Functions|[TMI](https://ieeexplore.ieee.org/document/9229101)|
|20200929|Stefan Gerl|A **Distance-Based Loss** for Smooth and Continuous Skin Layer Segmentation in Optoacoustic Images|[MICCAI 2020](https://link.springer.com/chapter/10.1007%2F978-3-030-59725-2_30)|
|20200821|Nick Byrne|A persistent homology-based **topological loss** function for multi-class CNN segmentation of cardiac MRI [arxiv](https://arxiv.org/abs/2008.09585)|STACOM|
|20200720|Boris Shirokikh|**Universal Loss Reweighting** to Balance Lesion Size Inequality in 3D Medical Image Segmentation [arxiv](https://arxiv.org/abs/2007.10033) [(pytorch)](https://arxiv.org/abs/2007.10033)|MICCAI 2020|
|20200708|Gonglei Shi|**Marginal loss and exclusion loss** for partially supervised multi-organ segmentation [(arXiv)](https://arxiv.org/abs/2007.03868)|MedIA|
|20200706|Yuan Lan|An **Elastic Interaction-Based Loss** Function for Medical Image Segmentation [(pytorch)](https://github.com/charrywhite/elastic_interaction_based_loss) [(arXiv)](https://arxiv.org/abs/2007.02663)|MICCAI 2020|
|20200615|Tom Eelbode|Optimization for Medical Image Segmentation: Theory and Practice when evaluating with Dice Score or Jaccard Index|[TMI](https://ieeexplore.ieee.org/document/9116807)|
|20200605|Guotai Wang|**Noise-robust Dice loss:** A Noise-robust Framework for Automatic Segmentation of COVID-19 Pneumonia Lesions from CT Images [(pytorch)](https://github.com/HiLab-git/COPLE-Net)|[TMI](https://ieeexplore.ieee.org/document/9109297)|
|202004|J. H. Moltz|**Contour Dice coefficient (CDC) Loss:** Learning a Loss Function for Segmentation: A Feasibility Study|[ISBI](https://ieeexplore.ieee.org/abstract/document/9098557)|
|201912|Yuan Xue|Shape-Aware Organ Segmentation by Predicting Signed Distance Maps [(arxiv)](https://arxiv.org/abs/1912.03849) [(pytorch)](https://github.com/JunMa11/SegWithDistMap/blob/master/code/train_LA_AAAISDF.py)|AAAI 2020|
|201912|Xiaoling Hu|**Topology-Preserving** Deep Image Segmentation [(paper)](https://papers.nips.cc/paper/8803-topology-preserving-deep-image-segmentation.pdf) [(pytorch)](https://github.com/HuXiaoling/TopoLoss)|[NeurIPS](https://papers.nips.cc/paper/8803-topology-preserving-deep-image-segmentation)|
|201910|Shuai Zhao|Region Mutual Information Loss for Semantic Segmentation [(paper)](https://papers.nips.cc/paper/9291-region-mutual-information-loss-for-semantic-segmentation) [(pytorch)](https://github.com/ZJULearning/RMI)|[NeurIPS 2019](https://papers.nips.cc/paper/9291-region-mutual-information-loss-for-semantic-segmentation)|
|201910|Shuai Zhao|Correlation Maximized Structural Similarity Loss for Semantic Segmentation [(paper)](https://arxiv.org/abs/1910.08711)|arxiv|
|201908|Pierre-AntoineGanaye|Removing Segmentation Inconsistencies with Semi-Supervised Non-Adjacency Constraint [(paper)](https://www.sciencedirect.com/science/article/pii/S1361841519300866?dgcid=raven_sd_aip_email) [(official pytorch)](https://github.com/trypag/NonAdjLoss)|[Medical Image Analysis](https://www.sciencedirect.com/science/article/pii/S1361841519300866?dgcid=raven_sd_aip_email)|
|201906|Xu Chen|Learning **Active Contour Models** for Medical Image Segmentation [(paper)](http://openaccess.thecvf.com/content_CVPR_2019/papers/Chen_Learning_Active_Contour_Models_for_Medical_Image_Segmentation_CVPR_2019_paper.pdf) [(official-keras)](https://github.com/xuuuuuuchen/Active-Contour-Loss/blob/master/Active-Contour-Loss.py)|CVPR 2019|
|20190422|Davood Karimi|Reducing the **Hausdorff Distance** in Medical Image Segmentation with Convolutional Neural Networks [(paper)](https://arxiv.org/pdf/1904.10030v1.pdf) [(pytorch)](https://github.com/JunMa11/SegWithDistMap/blob/5a67153bc730eb82de396ef63f57594f558e23cd/code/train_LA_HD.py#L106)|[TMI 201907](https://ieeexplore.ieee.org/document/8767031)|
|20190417|Francesco Caliva|**Distance Map Loss** Penalty Term for Semantic Segmentation [(paper)](https://openreview.net/forum?id=B1eIcvS45V)|[MIDL 2019](http://2019.midl.io/)|
|20190411|Su Yang|Major Vessel Segmentation on X-ray Coronary Angiography using Deep Networks with a Novel **Penalty Loss Function** [(paper)](https://openreview.net/forum?id=H1lTh8unKN)|[MIDL 2019](http://2019.midl.io/)|
|20190405|Boah Kim|Multiphase **Level-Set Loss** for Semi-Supervised and Unsupervised Segmentation with Deep Learning [(paper)](https://arxiv.org/pdf/1904.02872.pdf)|arxiv|
|201901|[Seyed Raein Hashemi](https://scholar.google.ca/citations?user=4VEP0fsAAAAJ&hl=en&oi=sra)|**Asymmetric Loss** Functions and Deep Densely Connected Networks for Highly Imbalanced Medical Image Segmentation: Application to Multiple Sclerosis Lesion Detection [(paper)](https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8573779)|IEEE Access|
|201812|[Hoel Kervadec](https://scholar.google.ca/citations?user=yeFGhfgAAAAJ&hl=zh-CN&oi=sra)|**Boundary loss** for highly unbalanced segmentation [(paper)](https://arxiv.org/pdf/1812.07032.pdf), [(pytorch 1.0)](https://github.com/LIVIAETS/surface-loss)|[MIDL 2019](http://2019.midl.io/)|
|201810|[Nabila Abraham](https://scholar.google.ca/citations?user=OOvooSMAAAAJ&hl=zh-CN&oi=sra)|A Novel **Focal Tversky loss** function with improved Attention U-Net for lesion segmentation [(paper)](https://arxiv.org/pdf/1810.07842.pdf) [(keras)](https://github.com/nabsabraham/focal-tversky-unet)|[ISBI 2019](https://biomedicalimaging.org/2019/)|
|201809|[Fabian Isensee](https://scholar.google.com/citations?user=PjerEe4AAAAJ&hl=en)|**CE+Dice:** nnU-Net: Self-adapting Framework for U-Net-Based Medical Image Segmentation [(paper)](https://arxiv.org/abs/1809.10486)|arxiv|
|20180831|[Ken C. L. Wong](https://scholar.google.ca/citations?hl=zh-CN&user=XjnODToAAAAJ&view_op=list_works&sortby=pubdate)|3D Segmentation with **Exponential Logarithmic Loss** for Highly Unbalanced Object Sizes [(paper)](https://arxiv.org/abs/1809.00076)|MICCAI 2018|
|20180815|[Wentao Zhu](https://www.ics.uci.edu/~wentaoz1/)|**Dice+Focal:** AnatomyNet: Deep Learning for Fast and Fully Automated Whole-volume Segmentation of Head and Neck Anatomy [(arxiv)](https://arxiv.org/abs/1808.05238) [(pytorch)](https://github.com/wentaozhu/AnatomyNet-for-anatomical-segmentation)|[Medical Physics](https://aapm.onlinelibrary.wiley.com/doi/full/10.1002/mp.13300)|
|201806|[Javier Ribera](https://scholar.google.ca/citations?user=TAaovakAAAAJ&hl=zh-CN&oi=sra)|**Weighted Hausdorff Distance:** Locating Objects Without Bounding Boxes [(paper)](https://arxiv.org/abs/1806.07564), [(pytorch)](https://github.com/HaipengXiong/weighted-hausdorff-loss)|CVPR 2019|
|201805|Saeid Asgari Taghanaki|**Combo Loss:** Handling Input and Output Imbalance in Multi-Organ Segmentation [(arxiv)](https://arxiv.org/pdf/1805.02798.pdf) [(keras)](https://github.com/asgsaeid/ComboLoss/blob/master/combo_loss.py)|[Computerized Medical Imaging and Graphics](https://www.sciencedirect.com/science/article/abs/pii/S0895611118305688)|
|201709|[S M Masudur Rahman AL ARIF](https://scholar.google.ca/citations?user=6bgRPC8AAAAJ&hl=en&oi=sra)|**Shape-aware** deep convolutional neural network for vertebrae segmentation [(paper)](http://www.gregslabaugh.net/publications/ArifMSKI-MICCAI2017.pdf)|[MICCAI 2017 Workshop](https://link.springer.com/chapter/10.1007/978-3-319-74113-0_2)|
|201708|[Tsung-Yi Lin](https://scholar.google.ca/citations?user=_BPdgV0AAAAJ&hl=zh-CN&oi=sra)|**Focal Loss** for Dense Object Detection [(paper)](https://arxiv.org/abs/1708.02002), [(code)](https://github.com/facebookresearch/Detectron)|ICCV, TPAMI|
|20170711|[Carole Sudre](https://scholar.google.ca/citations?user=14GfvB4AAAAJ&hl=zh-CN&oi=sra)|**Generalised Dice** overlap as a deep learning loss function for highly unbalanced segmentations [(paper)](https://arxiv.org/abs/1707.03237)|DLMIA 2017|
|20170703|[Lucas Fidon](https://scholar.google.ca/citations?user=GORojioAAAAJ&hl=zh-CN&oi=sra)|**Generalised Wasserstein Dice** Score for Imbalanced Multi-class Segmentation using Holistic Convolutional Networks [(paper)](https://arxiv.org/abs/1707.00478)|MICCAI 2017 BrainLes|
|201705|[Maxim Berman](https://scholar.google.ca/citations?user=RoOng2wAAAAJ&hl=zh-CN&oi=sra)|The **Lovász-Softmax loss:** A tractable surrogate for the optimization of the intersection-over-union measure in neural networks [(paper)](https://arxiv.org/abs/1705.08790), [(code)](https://github.com/bermanmaxim/LovaszSoftmax)|CVPR 2018|
|201701|[Seyed Sadegh Mohseni Salehi](https://scholar.google.ca/citations?user=hTWINokAAAAJ&hl=zh-CN&oi=sra)|**Tversky loss** function for image segmentation using 3D fully convolutional deep networks [(paper)](https://arxiv.org/abs/1706.05721)|MICCAI 2017 MLMI|
|201612|[Md Atiqur Rahman](https://scholar.google.ca/citations?user=tLPerVUAAAAJ&hl=zh-CN&oi=sra)|Optimizing **Intersection-Over-Union** in Deep Neural Networks for Image Segmentation [(paper)](https://www.cs.umanitoba.ca/~ywang/papers/isvc16.pdf)|[2016 International Symposium on Visual Computing](https://link.springer.com/chapter/10.1007/978-3-319-50835-1_22)|
|201608|[Michal Drozdzal](https://scholar.google.es/citations?user=XK_ktwQAAAAJ&hl=en)|**"Dice Loss (without square)"** The Importance of Skip Connections in Biomedical Image Segmentation [(arxiv)](https://arxiv.org/abs/1608.04117)|[DLMIA 2016](https://link.springer.com/chapter/10.1007/978-3-319-46976-8_19)|
|201606|[Fausto Milletari](https://faustomilletari.github.io/)|**"Dice Loss (with square)"** V-net: Fully convolutional neural networks for volumetric medical image segmentation [(arxiv)](https://arxiv.org/abs/1606.04797), [(caffe code)](https://github.com/faustomilletari/VNet)|International Conference on 3D Vision|
|201605|Zifeng Wu|**TopK loss** Bridging Category-level and Instance-level Semantic Image Segmentation [(paper)](https://arxiv.org/abs/1605.06885)|arxiv|
|201511|[Tom Brosch](https://scholar.google.ca/citations?user=KChq7WIAAAAJ&hl=zh-CN&oi=sra)|**"Sensitivity-Specifity loss"** Deep Convolutional Encoder Networks for Multiple Sclerosis Lesion Segmentation [(paper)](http://www.rogertam.ca/Brosch_MICCAI_2015.pdf) [(code)](https://github.com/NifTK/NiftyNet/blob/df0f86733357fdc92bbc191c8fec0dcf49aa5499/niftynet/layer/loss_segmentation.py#L392)|[MICCAI 2015](https://link.springer.com/chapter/10.1007/978-3-319-24574-4_1)|
|201505|[Olaf Ronneberger](https://scholar.google.ca/citations?user=7jrO1NwAAAAJ&hl=zh-CN&oi=sra)|**"Weighted cross entropy"** U-Net: Convolutional Networks for Biomedical Image Segmentation [(paper)](https://arxiv.org/abs/1505.04597)|[MICCAI 2015](https://link.springer.com/chapter/10.1007/978-3-319-24574-4_28)|
|201309|[Gabriela Csurka](https://scholar.google.ca/citations?user=PXm1lPAAAAAJ&hl=zh-CN&oi=sra)|What is a good evaluation measure for semantic segmentation? [(paper)](http://www.bmva.org/bmvc/2013/Papers/paper0032/paper0032.pdf)|BMVA 2013|

> Most of the corresponding tensorflow code can be found [here](https://github.com/NifTK/NiftyNet/blob/dev/niftynet/layer/loss_segmentation.py).