Asymmetric Contour Uncertainty Estimation for Medical Image Segmentation

Olivier Bernard, Thierry Judge, Pierre-marc Jodoin, Agisilaos Chartsias, Alberto Gomez

Abstract: Aleatoric uncertainty estimation is a critical step in medical image segmentation. Most techniques for estimating aleatoric uncertainty for segmentation purposes assume a Gaussian distribution over the neural network's logit value modeling the uncertainty in the presence of a class at that location. However, in many cases of segmentation, such as heart ultrasound or chest X-ray segmentation, there is no uncertainty about the presence of a specific structure but rather about the precise outline of that structure. For this reason, we propose to explicitly model the location uncertainty by reframing the usual pixel-by-pixel segmentation task into a contour regression problem. This allows for modeling the uncertainty of contour points using a more appropriate multivariate distribution. Also, since countour uncertainly is often anisotropic, we use a multivariate skewed Gaussian distribution. In addition to being directly interpretable, our uncertainty estimation method outperforms previous methods on three datasets using two different image modalities