Retinal vessel segmentation based on Fully Convolutional Neural Networks

作者：

Highlights：

• A multi-scale Fully Convolutional Neural Network for retinal vessel segmentation.

• Rotations as the basis of a joint strategy for data augmentation and prediction.

• The Stationary Wavelet Transform as a way to stack new channels into the input.

摘要

•A multi-scale Fully Convolutional Neural Network for retinal vessel segmentation.•Rotations as the basis of a joint strategy for data augmentation and prediction.•The Stationary Wavelet Transform as a way to stack new channels into the input.

论文关键词：Fully Convolutional Neural Network,Stationary Wavelet Transform,Retinal fundus image,Vessel segmentation,Deep learning

论文评审过程：Received 10 January 2018, Revised 16 May 2018, Accepted 14 June 2018, Available online 18 June 2018, Version of Record 27 July 2018.

论文官网地址：https://doi.org/10.1016/j.eswa.2018.06.034