2021
DOI: 10.1186/s12880-021-00694-1
|View full text |Cite
|
Sign up to set email alerts
|

Accurate pancreas segmentation using multi-level pyramidal pooling residual U-Net with adversarial mechanism

Abstract: Background A novel multi-level pyramidal pooling residual U-Net with adversarial mechanism was proposed for organ segmentation from medical imaging, and was conducted on the challenging NIH Pancreas-CT dataset. Methods The 82 pancreatic contrast-enhanced abdominal CT volumes were split via four-fold cross validation to test the model performance. In order to achieve accurate segmentation, we firstly involved residual learning into an adversarial U-… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
7
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
7
1
1

Relationship

0
9

Authors

Journals

citations
Cited by 18 publications
(7 citation statements)
references
References 18 publications
(9 reference statements)
0
7
0
Order By: Relevance
“…Adversarial loss combined with multi-level pyramid pooling modules [39] optimises the multi-scale representation of features and strengthens the model's discriminative ability for different structures and textures. The merit lies in its excellent performance in handling class imbalance and irregular target segmentation, which may lead to model overfitting [39].…”
Section: Convolution Block Modificationsmentioning
confidence: 99%
See 1 more Smart Citation
“…Adversarial loss combined with multi-level pyramid pooling modules [39] optimises the multi-scale representation of features and strengthens the model's discriminative ability for different structures and textures. The merit lies in its excellent performance in handling class imbalance and irregular target segmentation, which may lead to model overfitting [39].…”
Section: Convolution Block Modificationsmentioning
confidence: 99%
“…Multi-Level Pyramid Pooling Residual U-Net improves performance by incorporating a Multi-Level Pyramid Pooling (MLPP) module [39]. MLPP employs pooling at various scales to capture multi-scale information, combined with residual learning and adversarial training.…”
Section: E Alternative Pooling Strategymentioning
confidence: 99%
“…Li et al (Li et al, 2021) aimed to segmentation of the pancreas using abdominal CT images. The NIH data set was used in this study.…”
Section: Related Workmentioning
confidence: 99%
“…Therefore, such a deficiency in capturing multi-scale information yields a performance degradation in the segmentation of complex structures, with variation in shapes and scales. Different methods have been proposed to solve the problem of the restricted receptive field of regular CNNs in recent years [11,22]. Wang et al [29] extended the self-attention concept into the spatial domain to model non-local properties of images by devising a non-local module that can be easily integrated into existing network designs.…”
Section: Introductionmentioning
confidence: 99%