Collection of loss functions in medical image segmentation
- Using multi task learning as regularization
- Dice loss and its variants
- Cross-Entropy loss and its variants
- Tversky loss and its variants
- Adversarial loss
- Auto-encoder loss
- Micellaneous: Overlapping loss, Threshold map loss
2016
- V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation, Fausto Milletari et al, Arxiv. pdf
2017
- Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentations, Carole H. Sudre et al, Arxiv. pdf
2018
- Densely Deep Supervised Networks with Threshold Loss for Cancer Detection in Automated Breast Ultrasound, Na Wang et al, MICCAI. pdf
- A Novel Focal Tversky Loss Function with Improved Attentation Unet for Lesion Segmentation, Nabila Abraham et al, Arxiv. pdf
- AnatomyNet: Deep Learning for Fast and Fully Automated Whole-volume Segmentation of Head and Neck Anatomy, Wentao Zhu et al, Arxiv. pdf
- 3D Segmentation with Exponential Logarithmic Loss for Highly Unbalanced Object Sizes, Ken C. L. Wong et al, Arxiv. pdf
- BESNet: Boundary-Enhanced Segmentation of Cellsin Histopathological Images, Hirohisa Oda et al, Arxiv. pdf
- Deep Multi-Task and Task specific Feature Learning Network for Roubust Shape Preserved Organ Segmentation, Chaowei Tan et al, ISBI. pdf
2019
- A Distance Map Regularized CNN for Cardiac Cine MR Image Segmentation, Shusil Dangi et al, Arxiv. pdf
- Psi-Net: Shape and boundary aware joint multi-task deep network for medical image segmentation, Balamurali Murugesan et al, Arxiv. pdf
- Bottleneck Supervised U-Net for Pixel-wise Liver and Tumor Segmentation, Song LI et al, Arxiv. pdf
- Automatic Segmentation of Vestibular Schwannoma from T2-Weighted MRI by Deep Spatial Attention with Hardness-Weighted Loss, Guotai Wang et al, Arxiv. pdf