Privileged Modality Guided Network for Retinal Vessel Segmentation in Ultra-Wide-Field Images

Li, Xuefei1,2; Hao, Huaying1; Fu, Huazhu3; Zhang, Dan4; Chen, Da5; Qiao, Yuchuan6; Liu, Jiang7; Zhao, Yitian1; Zhang, Jiong1

10.1007/978-3-031-44013-7_9

2023

10th International Workshop on Ophthalmic Medical Image Analysis (OMIA)

0302-9743

1611-3349

978-3-031-44012-0

OPHTHALMIC MEDICAL IMAGE ANALYSIS, OMIA 2023

14096

OCT 12, 2023

null,Vancouver,CANADA

GEWERBESTRASSE 11, CHAM, CH-6330, SWITZERLAND

SPRINGER INTERNATIONAL PUBLISHING AG

Retinal vessel segmentation in ophthalmic images is an essential task to support the computer-aided diagnosis of eye-related diseases. As a non-invasive imaging technique, ultra-wide-field (UWF) fundus imaging provides a large field-of-view (FOV) of 200. with full coverage of the retinal territory, making it a suitable modality for vessel analysis. However, imaging the large FOV may result in low-contrast vascular details and background artifacts, which pose challenges to the accurate segmentation of retinal microvasculature. To address these issues, a privileged modality guided multi-scale location-aware fusion network is proposed for vessel segmentation in UWF images. We first perform style transfer on the UWF images to generate the corresponding FFA image with higher contrast. Afterwards, we employ cross-modal coherence loss to segment the vessels guided by the FFA image. Additionally, a multi-scale location-aware fusion module is proposed and embedded into the segmentation network for reducing the boundary artifacts. Finally, experiments are performed on a dedicated UWF dataset, and the evaluation results demonstrate that our method achieves competitive vessel segmentation performance with a Dice score of around 78.13%. This indicates that our method is potentially valuable for subsequent vessel analysis to support disease diangosis.

UWF Vessel segmentation Privileged modality Location-aware

其他

英语

EI ; CPCI-S

Computer Science ; Engineering ; Ophthalmology

Computer Science, Artificial Intelligence ; Computer Science, Theory & Methods ; Engineering, Biomedical ; Ophthalmology

WOS:001116062200009

Web of Science

1.Institute of Biomedical Engineering, Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo, China
2.University of Chinese Academy of Sciences, Beijing, China
3.Institute of High Performance Computing, A*STAR, Singapore, Singapore
4.School of Cyber Science and Engineering, Ningbo University of Technology, Ningbo, China
5.Shandong Artificial Intelligence Institute, Jinan, China
6.Institute of Science and Technology for Brain-Inspired Intelligence, Fudan University, Shanghai, China
7.Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China

Li, Xuefei,Hao, Huaying,Fu, Huazhu,et al. Privileged Modality Guided Network for Retinal Vessel Segmentation in Ultra-Wide-Field Images[C]. GEWERBESTRASSE 11, CHAM, CH-6330, SWITZERLAND:SPRINGER INTERNATIONAL PUBLISHING AG,2023.