Boundary loss for highly unbalanced segmentation

By Hoel Kervadec, Jihene Bouchtiba, Christian Desrosiers, Eric Granger, Jose Dolz, Ismail Ben Ayed. International Conference on Medical Imaging with Deep Learning, 2018

736 citations recorded by the source corpus.

Widely used loss functions for CNN segmentation, e.g., Dice or cross-entropy, are based on integrals over the segmentation regions. Unfortunately, for highly unbalanced segmentations, such regional summations have values that differ by several orders of magnitude across classes, which affects training performance and stability. We propose a boundary loss, which takes the form of a distance metric on the space of contours, not regions. This can mitigate the difficulties of highly unbalanced problems because it uses integrals over the interface between regions instead of unbalanced integrals over the regions. Furthermore, a boundary loss complements regional information. Inspired by graph-based optimization techniques for computing active-contour flows, we express a non-symmetric L2 distance on the space of contours as a regional integral, which avoids completely local differential computations involving contour points. This yields a boundary loss expressed with the regional softmax probability outputs of the network, which can be easily combined with standard regional losses and implemented with any existing deep network architecture for N-D segmentation. We report comprehensive evaluations and comparisons on different unbalanced problems, showing that our boundary loss can yield significant increases in performances while improving training stability. Our code is publicly available1.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Neurology

View Paper: https://www.semanticscholar.org/paper/a1e392a596bc2f34fe34dddf79ee11c20fe5d00e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Neurology

Citations in the corpus

U-Net and Its Variants for Medical Image Segmentation: A Review of Theory and Applications (2020, background): “Lastly, boundary loss is a family of loss functions that aim to minimize the distance between the ground truth and the segmentation results on a regional basis [365].”
Embracing Imperfect Datasets: A Review of Deep Learning Solutions for Medical Image Segmentation (2019, methodology): “Inspired by the optimization technique for computing gradient flows of curve evolution, Kervadec et al. (2019) introduce a non-symmetric L2 loss to regularize boundary deviation of the segmentation mask S relative to the ground truth G,”
Unified Focal loss: Generalising Dice and cross entropy-based losses to handle class imbalanced medical image segmentation (2021, background): “Broadly, loss functions used in image segmentation may be classified into distribution-based losses (such as the cross entropy loss), region-based losses (such as Dice loss), boundary-based losses (such as the boundary loss) (Kervadec et al., 2019), and more recently compound losses.”
Medical Image Segmentation Using Deep Learning: A Survey (2020, methodology): “[95] proposed a new boundary loss used for brain lesion segmentation.”