Uncertainty-Guided Domain Alignment for Layer Segmentation in OCT Images

TL;DR

This paper introduces an uncertainty-guided domain adaptation approach for layer segmentation in OCT images, improving robustness across different equipment by leveraging uncertainty-aware loss, curriculum transfer, and adversarial learning.

Contribution

It proposes a novel uncertainty-guided domain alignment framework combining a new loss, curriculum transfer, and adversarial learning for OCT segmentation.

Findings

Significant segmentation accuracy improvements over baseline models.

Effective transfer of knowledge across different OCT datasets.

Robustness to equipment-induced appearance variations.

Abstract

Automatic and accurate segmentation for retinal and choroidal layers of Optical Coherence Tomography (OCT) is crucial for detection of various ocular diseases. However, because of the variations in different equipments, OCT data obtained from different manufacturers might encounter appearance discrepancy, which could lead to performance fluctuation to a deep neural network. In this paper, we propose an uncertainty-guided domain alignment method to aim at alleviating this problem to transfer discriminative knowledge across distinct domains. We disign a novel uncertainty-guided cross-entropy loss for boosting the performance over areas with high uncertainty. An uncertainty-guided curriculum transfer strategy is developed for the self-training (ST), which regards uncertainty as efficient and effective guidance to optimize the learning process in target domain. Adversarial learning with feature recalibration module (FRM) is applied to transfer informative knowledge from the domain feature spaces adaptively. The experiments on two OCT datasets show that the proposed methods can obtain significant segmentation improvements compared with the baseline models.