Generalised Dice Overlap as a Deep Learning Loss Function for Highly Unbalanced Segmentations
Top Cited Papers
- 9 September 2017
- book chapter
- conference paper
- Published by Springer Science and Business Media LLC in Lecture Notes in Computer Science
- Vol. 2017, 240-248
- https://doi.org/10.1007/978-3-319-67558-9_28
Abstract
Deep-learning has proved in recent years to be a powerful tool for image analysis and is now widely used to segment both 2D and 3D medical images. Deep-learning segmentation frameworks rely not only on the choice of network architecture but also on the choice of loss function. When the segmentation process targets rare observations, a severe class imbalance is likely to occur between candidate labels, thus resulting in sub-optimal performance. In order to mitigate this issue, strategies such as the weighted cross-entropy function, the sensitivity function or the Dice loss function, have been proposed. In this work, we investigate the behavior of these loss functions and their sensitivity to learning rate tuning in the presence of different rates of label imbalance across 2D and 3D segmentation tasks. We also propose to use the class re-balancing properties of the Generalized Dice overlap, a known metric for segmentation assessment, as a robust and accurate deep-learning loss function for unbalanced tasks.Keywords
Other Versions
This publication has 9 references indexed in Scilit:
- On the Compactness, Efficiency, and Representation of 3D Convolutional Networks: Brain Parcellation as a Pretext TaskLecture Notes in Computer Science, 2017
- Efficient multi-scale 3D CNN with fully connected CRF for accurate brain lesion segmentationMedical Image Analysis, 2017
- Brain tumor segmentation with Deep Neural NetworksMedical Image Analysis, 2017
- V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image SegmentationPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2016
- Deep Convolutional Encoder Networks for Multiple Sclerosis Lesion SegmentationPublished by Springer Science and Business Media LLC ,2015
- U-Net: Convolutional Networks for Biomedical Image SegmentationPublished by Springer Science and Business Media LLC ,2015
- 3D Deep Learning for Efficient and Robust Landmark Detection in Volumetric DataPublished by Springer Science and Business Media LLC ,2015
- The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS)IEEE Transactions on Medical Imaging, 2014
- Generalized Overlap Measures for Evaluation and Validation in Medical Image AnalysisIEEE Transactions on Medical Imaging, 2006