ErrorNet: Learning error representations from limited data to improve vascular segmentation

TL;DR

ErrorNet is a novel segmentation framework that learns to correct mistakes in medical image segmentation by injecting and predicting errors, improving accuracy especially with limited data and domain shifts without needing domain-specific tuning.

Contribution

This paper introduces ErrorNet, a new error correction framework for medical image segmentation that does not require domain-specific tuning or target domain data, addressing small sample size and domain shift issues.

Findings

ErrorNet outperforms baseline segmentation models.

ErrorNet surpasses CRF-based post processing.

ErrorNet shows greater improvements with limited data and domain shifts.

Abstract

Deep convolutional neural networks have proved effective in segmenting lesions and anatomies in various medical imaging modalities. However, in the presence of small sample size and domain shift problems, these models often produce masks with non-intuitive segmentation mistakes. In this paper, we propose a segmentation framework called ErrorNet, which learns to correct these segmentation mistakes through the repeated process of injecting systematic segmentation errors to the segmentation result based on a learned shape prior, followed by attempting to predict the injected error. During inference, ErrorNet corrects the segmentation mistakes by adding the predicted error map to the initial segmentation result. ErrorNet has advantages over alternatives based on domain adaptation or CRF-based post processing, because it requires neither domain-specific parameter tuning nor any data from the target domains. We have evaluated ErrorNet using five public datasets for the task of retinal vessel segmentation. The selected datasets differ in size and patient population, allowing us to evaluate the effectiveness of ErrorNet in handling small sample size and domain shift problems. Our experiments demonstrate that ErrorNet outperforms a base segmentation model, a CRF-based post processing scheme, and a domain adaptation method, with a greater performance gain in the presence of the aforementioned dataset limitations.