Stack-U-Net: Refinement Network for Image Segmentation on the Example of Optic Disc and Cup

TL;DR

This paper introduces Stack-U-Net, a cascade refinement network based on U-Net architecture, achieving superior optic disc and cup segmentation quality for glaucoma detection without requiring larger datasets.

Contribution

The paper presents a novel cascade U-Net-based model that enhances segmentation accuracy through iterative refinement, outperforming existing methods on multiple datasets.

Findings

Achieved higher segmentation accuracy than state-of-the-art methods.

Effective on multiple publicly available datasets and private data.

No need for larger datasets to improve performance.

Abstract

In this work, we propose a special cascade network for image segmentation, which is based on the U-Net networks as building blocks and the idea of the iterative refinement. The model was mainly applied to achieve higher recognition quality for the task of finding borders of the optic disc and cup, which are relevant to the presence of glaucoma. Compared to a single U-Net and the state-of-the-art methods for the investigated tasks, very high segmentation quality has been achieved without a need for increasing the volume of datasets. Our experiments include comparison with the best-known methods on publicly available databases DRIONS-DB, RIM-ONE v.3, DRISHTI-GS, and evaluation on a private data set collected in collaboration with University of California San Francisco Medical School. The analysis of the architecture details is presented, and it is argued that the model can be employed for a broad scope of image segmentation problems of similar nature.