Reducing the Hausdorff Distance in Medical Image Segmentation With Convolutional Neural Networks

By Davood Karimi, Septimiu E. Salcudean. IEEE Transactions on Medical Imaging, 2019

534 citations recorded by the source corpus.

The Hausdorff Distance (HD) is widely used in evaluating medical image segmentation methods. However, the existing segmentation methods do not attempt to reduce HD directly. In this paper, we present novel loss functions for training convolutional neural network (CNN)-based segmentation methods with the goal of reducing HD directly. We propose three methods to estimate HD from the segmentation probability map produced by a CNN. One method makes use of the distance transform of the segmentation boundary. Another method is based on applying morphological erosion on the difference between the true and estimated segmentation maps. The third method works by applying circular/spherical convolution kernels of different radii on the segmentation probability maps. Based on these three methods for estimating HD, we suggest three loss functions that can be used for training to reduce HD. We use these loss functions to train CNNs for segmentation of the prostate, liver, and pancreas in ultrasound, magnetic resonance, and computed tomography images and compare the results with commonly-used loss functions. Our results show that the proposed loss functions can lead to approximately 18–45% reduction in HD without degrading other segmentation performance criteria such as the Dice similarity coefficient. The proposed loss functions can be used for training medical image segmentation methods in order to reduce the large segmentation errors.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/05143a909ce9db357c2a4001d40fe0b29785d613

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

Citations in the corpus

Medical Image Segmentation Using Deep Learning: A Survey (2020, background): “[100] proposed a loss function based on Hausdorff distance (HD).”
Boundary loss for highly unbalanced segmentation (2018, methodology): “Hausdorff loss ( Karimi and Salcudean, 2019 ) This closely re-lated loss is also designed to minimize some distance between the two boundaries, but through a different path.”
Sharp U-Net: Depthwise Convolutional Network for Biomedical Image Segmentation (2021, background): “It is worth mentioning that other loss functions such as the focal Tversky loss [25] and distance-based losses [26,27] can also be used for training.”
A survey of loss functions for semantic segmentation (2020, background): “Hausdorff Distance between point sets X and Y [18]”