Dual encoder–decoder-based deep polyp segmentation network for colonoscopy images
Abstract Detection of colorectal polyps through colonoscopy is an essential practice in prevention of colorectal cancers. However, the method itself is labor intensive and is subject to human error. With the advent of deep learning-based methodologies, and specifically convolutional neural networks,...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-01-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-023-28530-2 |
| _version_ | 1797945990078005248 |
|---|---|
| author | John Lewis Young-Jin Cha Jongho Kim |
| author_facet | John Lewis Young-Jin Cha Jongho Kim |
| author_sort | John Lewis |
| collection | DOAJ |
| description | Abstract Detection of colorectal polyps through colonoscopy is an essential practice in prevention of colorectal cancers. However, the method itself is labor intensive and is subject to human error. With the advent of deep learning-based methodologies, and specifically convolutional neural networks, an opportunity to improve upon the prognosis of potential patients suffering with colorectal cancer has appeared with automated detection and segmentation of polyps. Polyp segmentation is subject to a number of problems such as model overfitting and generalization, poor definition of boundary pixels, as well as the model’s ability to capture the practical range in textures, sizes, and colors. In an effort to address these challenges, we propose a dual encoder–decoder solution named Polyp Segmentation Network (PSNet). Both the dual encoder and decoder were developed by the comprehensive combination of a variety of deep learning modules, including the PS encoder, transformer encoder, PS decoder, enhanced dilated transformer decoder, partial decoder, and merge module. PSNet outperforms state-of-the-art results through an extensive comparative study against 5 existing polyp datasets with respect to both mDice and mIoU at 0.863 and 0.797, respectively. With our new modified polyp dataset we obtain an mDice and mIoU of 0.941 and 0.897 respectively. |
| first_indexed | 2024-04-10T21:03:47Z |
| format | Article |
| id | doaj.art-f5c2c4a6cb914ec983ba94a47a9f82ba |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-10T21:03:47Z |
| publishDate | 2023-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-f5c2c4a6cb914ec983ba94a47a9f82ba2023-01-22T12:12:19ZengNature PortfolioScientific Reports2045-23222023-01-0113111210.1038/s41598-023-28530-2Dual encoder–decoder-based deep polyp segmentation network for colonoscopy imagesJohn Lewis0Young-Jin Cha1Jongho Kim2Department of Civil Engineering, University of ManitobaDepartment of Civil Engineering, University of ManitobaDepartment of Radiology, Max Rady College of Medicine, University of ManitobaAbstract Detection of colorectal polyps through colonoscopy is an essential practice in prevention of colorectal cancers. However, the method itself is labor intensive and is subject to human error. With the advent of deep learning-based methodologies, and specifically convolutional neural networks, an opportunity to improve upon the prognosis of potential patients suffering with colorectal cancer has appeared with automated detection and segmentation of polyps. Polyp segmentation is subject to a number of problems such as model overfitting and generalization, poor definition of boundary pixels, as well as the model’s ability to capture the practical range in textures, sizes, and colors. In an effort to address these challenges, we propose a dual encoder–decoder solution named Polyp Segmentation Network (PSNet). Both the dual encoder and decoder were developed by the comprehensive combination of a variety of deep learning modules, including the PS encoder, transformer encoder, PS decoder, enhanced dilated transformer decoder, partial decoder, and merge module. PSNet outperforms state-of-the-art results through an extensive comparative study against 5 existing polyp datasets with respect to both mDice and mIoU at 0.863 and 0.797, respectively. With our new modified polyp dataset we obtain an mDice and mIoU of 0.941 and 0.897 respectively.https://doi.org/10.1038/s41598-023-28530-2 |
| spellingShingle | John Lewis Young-Jin Cha Jongho Kim Dual encoder–decoder-based deep polyp segmentation network for colonoscopy images Scientific Reports |
| title | Dual encoder–decoder-based deep polyp segmentation network for colonoscopy images |
| title_full | Dual encoder–decoder-based deep polyp segmentation network for colonoscopy images |
| title_fullStr | Dual encoder–decoder-based deep polyp segmentation network for colonoscopy images |
| title_full_unstemmed | Dual encoder–decoder-based deep polyp segmentation network for colonoscopy images |
| title_short | Dual encoder–decoder-based deep polyp segmentation network for colonoscopy images |
| title_sort | dual encoder decoder based deep polyp segmentation network for colonoscopy images |
| url | https://doi.org/10.1038/s41598-023-28530-2 |
| work_keys_str_mv | AT johnlewis dualencoderdecoderbaseddeeppolypsegmentationnetworkforcolonoscopyimages AT youngjincha dualencoderdecoderbaseddeeppolypsegmentationnetworkforcolonoscopyimages AT jonghokim dualencoderdecoderbaseddeeppolypsegmentationnetworkforcolonoscopyimages |