A multi-scale gated network for retinal hemorrhage detection

Retinal hemorrhage detection is of great significance for clinical diagnosis and disease control. However, most of the traditional methods need to obtain candidate lesions firstly, and then determine the true lesions. To address this problem, we propose an end-to-end multi-scale gated network (MGNet...

Full description

Saved in:
Bibliographic Details
Published in:Applied intelligence (Dordrecht, Netherlands) Netherlands), 2023-03, Vol.53 (5), p.5259-5273
Main Authors: Xia, Haiying, Rao, Zengyan, Zhou, Zuoshan
Format: Article
Language:English
Subjects:
Artificial Intelligence
Blood vessels
Candidates
Computer Science
Deep learning
Diabetes
Diabetic retinopathy
Disease control
Hemorrhage
Lesions
Machine learning
Machines
Manufacturing
Mechanical Engineering
Modules
Morphology
Processes
Support vector machines
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Retinal hemorrhage detection is of great significance for clinical diagnosis and disease control. However, most of the traditional methods need to obtain candidate lesions firstly, and then determine the true lesions. To address this problem, we propose an end-to-end multi-scale gated network (MGNet) to directly detect hemorrhage. Taken the U-Net as the backbone, we first add a skip connection gated module (SGate) to the skip connection to suppress useless information. Secondly, we propose a high-resolution and low-resolution multi-scale fusion module (HLMS) to improve the representation capacity of network through fusing information from the adjacent decoder layers. Furtherly, we propose a weighted Dice loss (W-Dice loss) to focus on the hard samples. Extensive experiments on two publicly available datasets: DIARETDB1 and IDRiD, demonstrate that the proposed MGNet achieves competitive results of hemorrhage detection compared with the state-of-the-art works.
ISSN:0924-669X
1573-7497
DOI:10.1007/s10489-022-03476-6