Table 2 Segmentation evaluation of our model against prior algorithms on low-contrast vessels in retinal images.

From: Exploring a multi-path U-net with probability distribution attention and cascade dilated convolution for precise retinal vessel segmentation in fundus images

Methods

Year

\(A_{cc}\)

SE

SP

\(F_{1}\) score

AUC

DUNet67

2019

0.9846

0.8858

0.9837

0.8889

0.9922

DCU-net68

2022

0.9845

0.8905

0.9923

0.8894

0.9934

SRV-GAN65

2022

0.9834

0.8991

0.9899

0.8923

0.9928

LUVS-Net66

2023

0.9844

0.8977

0.9911

0.8915

0.9932

ARDC-UNet69

2024

0.9846

0.8974

0.9929

0.8906

0.9936

ResMU-Net70

2024

0.9857

0.8943

0.9942

0.8959

0.9926

Ours

–

0.9869

0.8984

0.9925

0.9023

0.9932

Back to article page