2022
DOI: 10.1109/jbhi.2022.3188710
|View full text |Cite
|
Sign up to set email alerts
|

Full-Resolution Network and Dual-Threshold Iteration for Retinal Vessel and Coronary Angiograph Segmentation

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
23
0
4

Year Published

2023
2023
2024
2024

Publication Types

Select...
3
3
1

Relationship

0
7

Authors

Journals

citations
Cited by 91 publications
(27 citation statements)
references
References 31 publications
0
23
0
4
Order By: Relevance
“…The FR-UNet network, which was originally proposed for the detection of blood vessels in color fundus photographs, obtains the highest values for selected metrics in the case of OCT-reconstructed fundus. For this neural network, the impact of the dual-threshold iterative (DTI) algorithm [ 17 ] was also tested. The accuracy, sensitivity, F1 score, and AUC for FR-UNet are the highest (for the P3 reconstruction).…”
Section: Resultsmentioning
confidence: 99%
See 2 more Smart Citations
“…The FR-UNet network, which was originally proposed for the detection of blood vessels in color fundus photographs, obtains the highest values for selected metrics in the case of OCT-reconstructed fundus. For this neural network, the impact of the dual-threshold iterative (DTI) algorithm [ 17 ] was also tested. The accuracy, sensitivity, F1 score, and AUC for FR-UNet are the highest (for the P3 reconstruction).…”
Section: Resultsmentioning
confidence: 99%
“…The fifth neural network model considered in our application is a new approach called Full-Resolution and Dual-Threshold Iteration based on the UNet architecture [ 17 ]. It extends the original approach by horizontal and vertical expansion through a multiresolution convolution interactive mechanism.…”
Section: Methodsmentioning
confidence: 99%
See 1 more Smart Citation
“…We use the first annotation as the ground truth. Following [22], we use the first 20 images for training, and the remaining 8 images for testing.…”
Section: Dice Mioumentioning
confidence: 99%
“…Results (%) of Retinal Vessel Segmentation on CHASE DB1 dataset. The results of UNet, UNet++, Attention UNet, and FR-UNet are taken from[22]. All other results are averaged over five runs in our experimental setups.…”
mentioning
confidence: 99%