Uncertainty quantification in medical image segmentation with normalizing flows

Abstract

Medical image segmentation is inherently an ambiguous task due to factors such as partial volumes and variations in anatomical definitions. While in most cases the segmentation uncertainty is around the border of structures of interest, there can also be considerable inter-rater differences. The class of conditional variational autoencoders (cVAE) offers a principled approach to inferring distributions over plausible segmentations that are conditioned on input images. Segmentation uncertainty estimated from samples of such distributions can be more informative than using pixel level probability scores. In this work, we propose a novel conditional generative model that is based on conditional Normalizing Flow (cFlow). The basic idea is to increase the expressivity of the cVAE by introducing a cFlow transformation step after the encoder. This yields improved approximations of the latent posterior distribution, allowing the model to capture richer segmentation variations. With this we show that the quality and diversity of samples obtained from our conditional generative model is enhanced. Performance of our model, which we call cFlow Net, is evaluated on two medical imaging datasets demonstrating substantial improvements in both qualitative and quantitative measures when compared to a recent cVAE based model. 3

Uncertain-DeepSSM: From Images to Probabilistic Shape Models

“…Aleatoric uncertainty can be captured by placing a distribution over the model output. In image segmentation tasks, this has been achieved by sampling segmentations from an estimated posterior distribution [10,34] and using conditional normalizing flows [44] to infer a distribution of plausible segmentations conditioned on the input image. These efforts succeed in providing shape segmentation with aleatoric uncertainty measures, but do not provide a shape representation that can be readily used for population-level statistical analyses.…”