Automated segmentation of the opt ic disc from fundus images using an asymmetric deep learning network

PATTERN RECOGNITION

卷:112

Article

Wang, Lei^1,2  Gu, Juan¹  Chen, Yize¹  Liang, Yuanbo¹  Zhang, Weijie^3,4  Pu, Jiantao^3,4  Chen, Hao¹ 

[1] Wenzhou Med Univ, Sch Ophthalmol & Optometry, Eye Hosp, Wenzhou 325027, Peoples R China

[2] Southeast Univ, Minist Educ, Key Lab Comp Network & Informat Integrat, Nanjing 211189, Peoples R China

[3] Univ Pittsburgh, Dept Radiol, Pittsburgh, PA 15213 USA

[4] Univ Pittsburgh, Dept Bioengn, Pittsburgh, PA 15213 USA

关键词: Segmentation;    Colour fundus images;    Optic disc;    Deep learning;    U-Net;   

DOI  :  10.1016/j.patcog.2020.107810

来源: Elsevier

【 摘 要 】

Accurate segmentation of the optic disc (OD) regions from colour fundus images is a critical procedure for computer-aided diagnosis of glaucoma. We present a novel deep learning network to automatically identify the OD regions. On the basis of the classical U-Net framework, we define a unique sub-network and a decoding convolutional block. The sub-network is used to preserve important textures and facilitate their detections, while the decoding block is used to improve the contrast of the regions-of-interest with their background. We integrate these two components into the classical U-Net framework to improve the accuracy and reliability of segmenting the OD regions depicted on colour fundus images. We train and evaluate the developed network using three publicly available datasets (i.e., MESSIDOR, ORIGA, and REFUGE). The results on an independent testing set (n = 1,970 images) show a segmentation performance with an average Dice similarity coefficient (DSC), intersection over union (IOU), and Matthew's correlation coefficient (MCC) of 0.9377, 0.8854, and 0.9383 when trained on the global field-of-view images, respectively, and 0.9735, 0.9494, and 0.9594 when trained on the local disc region images. When compared with the other three classical networks (i.e., the U-Net, M-Net, and Deeplabv3) on the same testing datasets, the developed network demonstrates a relatively higher performance. (c) 2021 Elsevier Ltd. All rights reserved.

【 授权许可】

Free