The architecture of the SVNC-Net.
<div><p>Accurate measurement of spleen volume is essential for the diagnosis of splenomegaly. While Computed Tomography (CT) is among the most reliable imaging modalities for this task, manual segmentation of the spleen is labor-intensive and impractical for routine clinical workflows. A...
Uloženo v:
| Hlavní autor: | |
|---|---|
| Další autoři: | , , , , |
| Vydáno: |
2025
|
| Témata: | |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| _version_ | 1849927627474731008 |
|---|---|
| author | Mehmet Zahid Genc (22683526) |
| author2 | Yaser Dalveren (22683529) Ali Kara (690960) Mohammad Derawi (22683532) Jan Kubicek (170285) Marek Penhaker (13014797) |
| author2_role | author author author author author |
| author_facet | Mehmet Zahid Genc (22683526) Yaser Dalveren (22683529) Ali Kara (690960) Mohammad Derawi (22683532) Jan Kubicek (170285) Marek Penhaker (13014797) |
| author_role | author |
| dc.creator.none.fl_str_mv | Mehmet Zahid Genc (22683526) Yaser Dalveren (22683529) Ali Kara (690960) Mohammad Derawi (22683532) Jan Kubicek (170285) Marek Penhaker (13014797) |
| dc.date.none.fl_str_mv | 2025-11-25T18:35:17Z |
| dc.identifier.none.fl_str_mv | 10.1371/journal.pone.0332482.g001 |
| dc.relation.none.fl_str_mv | https://figshare.com/articles/figure/The_architecture_of_the_SVNC-Net_/30714409 |
| dc.rights.none.fl_str_mv | CC BY 4.0 info:eu-repo/semantics/openAccess |
| dc.subject.none.fl_str_mv | Medicine Space Science Environmental Sciences not elsewhere classified Biological Sciences not elsewhere classified Information Systems not elsewhere classified training compression techniques targeted architectural optimizations reliable imaging modalities promising results achieved neighborhood convolutional network model &# 8217 limited processing capabilities depthwise separable convolution high segmentation accuracy 3d organ segmentation routine clinical workflows memory demands make net builds upon entirely new architecture comparative analysis verified comparative analysis manual segmentation memory usage clinical deployment viable alternative time applications recent years purpose due ongoing efforts net variant net framework less suitable inference speed including upernet highly suitable first time findings contribute explore post edge devices constrained environments computed tomography 2d convolutions |
| dc.title.none.fl_str_mv | The architecture of the SVNC-Net. |
| dc.type.none.fl_str_mv | Image Figure info:eu-repo/semantics/publishedVersion image |
| description | <div><p>Accurate measurement of spleen volume is essential for the diagnosis of splenomegaly. While Computed Tomography (CT) is among the most reliable imaging modalities for this task, manual segmentation of the spleen is labor-intensive and impractical for routine clinical workflows. Automatic segmentation methods provide a more viable alternative for clinical deployment. In recent years, 3D Convolutional Neural Network (CNN) models have been widely used for this purpose due to their high segmentation accuracy. However, their computational and memory demands make them less suitable for real-time applications on edge devices with limited processing capabilities. To address these limitations, we introduce SVNC-Net (Spleen Volume and Neighborhood Convolutional Network) for efficient 3D spleen segmentation from CT scans. Rather than developing an entirely new architecture from scratch, SVNC-Net builds upon the U-Net framework with targeted architectural optimizations for efficiency. In SVNC-Net, each CT slice is processed independently using 2D convolutions. In its architecture, depthwise separable convolution is used to significantly reduce computational complexity and memory usage. To evaluate its performance and efficiency, a comparative analysis was conducted against well-known CNN-based models, including UPerNet, EMANet, CCNet, SegNet, and ShuffleNet. This evaluation was performed on two publicly available datasets used together for the first time in the literature. The promising results achieved from the comparative analysis verified that SVNC-Net is highly suitable for real-time applications and resource-constrained environments. Additionally, we explore post-training compression techniques such as pruning and quantization, which further enhance the model’s compactness and inference speed. These findings contribute to the ongoing efforts to develop efficient 2D deep learning models for 3D organ segmentation, particularly in resource-constrained clinical scenarios.</p></div> |
| eu_rights_str_mv | openAccess |
| id | Manara_0da8eb5cdd16516a297f3ce1d204af8d |
| identifier_str_mv | 10.1371/journal.pone.0332482.g001 |
| network_acronym_str | Manara |
| network_name_str | ManaraRepo |
| oai_identifier_str | oai:figshare.com:article/30714409 |
| publishDate | 2025 |
| repository.mail.fl_str_mv | |
| repository.name.fl_str_mv | |
| repository_id_str | |
| rights_invalid_str_mv | CC BY 4.0 |
| spelling | The architecture of the SVNC-Net.Mehmet Zahid Genc (22683526)Yaser Dalveren (22683529)Ali Kara (690960)Mohammad Derawi (22683532)Jan Kubicek (170285)Marek Penhaker (13014797)MedicineSpace ScienceEnvironmental Sciences not elsewhere classifiedBiological Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedtraining compression techniquestargeted architectural optimizationsreliable imaging modalitiespromising results achievedneighborhood convolutional networkmodel &# 8217limited processing capabilitiesdepthwise separable convolutionhigh segmentation accuracy3d organ segmentationroutine clinical workflowsmemory demands makenet builds uponentirely new architecturecomparative analysis verifiedcomparative analysismanual segmentationmemory usageclinical deploymentviable alternativetime applicationsrecent yearspurpose dueongoing effortsnet variantnet frameworkless suitableinference speedincluding upernethighly suitablefirst timefindings contributeexplore postedge devicesconstrained environmentscomputed tomography2d convolutions<div><p>Accurate measurement of spleen volume is essential for the diagnosis of splenomegaly. While Computed Tomography (CT) is among the most reliable imaging modalities for this task, manual segmentation of the spleen is labor-intensive and impractical for routine clinical workflows. Automatic segmentation methods provide a more viable alternative for clinical deployment. In recent years, 3D Convolutional Neural Network (CNN) models have been widely used for this purpose due to their high segmentation accuracy. However, their computational and memory demands make them less suitable for real-time applications on edge devices with limited processing capabilities. To address these limitations, we introduce SVNC-Net (Spleen Volume and Neighborhood Convolutional Network) for efficient 3D spleen segmentation from CT scans. Rather than developing an entirely new architecture from scratch, SVNC-Net builds upon the U-Net framework with targeted architectural optimizations for efficiency. In SVNC-Net, each CT slice is processed independently using 2D convolutions. In its architecture, depthwise separable convolution is used to significantly reduce computational complexity and memory usage. To evaluate its performance and efficiency, a comparative analysis was conducted against well-known CNN-based models, including UPerNet, EMANet, CCNet, SegNet, and ShuffleNet. This evaluation was performed on two publicly available datasets used together for the first time in the literature. The promising results achieved from the comparative analysis verified that SVNC-Net is highly suitable for real-time applications and resource-constrained environments. Additionally, we explore post-training compression techniques such as pruning and quantization, which further enhance the model’s compactness and inference speed. These findings contribute to the ongoing efforts to develop efficient 2D deep learning models for 3D organ segmentation, particularly in resource-constrained clinical scenarios.</p></div>2025-11-25T18:35:17ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pone.0332482.g001https://figshare.com/articles/figure/The_architecture_of_the_SVNC-Net_/30714409CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/307144092025-11-25T18:35:17Z |
| spellingShingle | The architecture of the SVNC-Net. Mehmet Zahid Genc (22683526) Medicine Space Science Environmental Sciences not elsewhere classified Biological Sciences not elsewhere classified Information Systems not elsewhere classified training compression techniques targeted architectural optimizations reliable imaging modalities promising results achieved neighborhood convolutional network model &# 8217 limited processing capabilities depthwise separable convolution high segmentation accuracy 3d organ segmentation routine clinical workflows memory demands make net builds upon entirely new architecture comparative analysis verified comparative analysis manual segmentation memory usage clinical deployment viable alternative time applications recent years purpose due ongoing efforts net variant net framework less suitable inference speed including upernet highly suitable first time findings contribute explore post edge devices constrained environments computed tomography 2d convolutions |
| status_str | publishedVersion |
| title | The architecture of the SVNC-Net. |
| title_full | The architecture of the SVNC-Net. |
| title_fullStr | The architecture of the SVNC-Net. |
| title_full_unstemmed | The architecture of the SVNC-Net. |
| title_short | The architecture of the SVNC-Net. |
| title_sort | The architecture of the SVNC-Net. |
| topic | Medicine Space Science Environmental Sciences not elsewhere classified Biological Sciences not elsewhere classified Information Systems not elsewhere classified training compression techniques targeted architectural optimizations reliable imaging modalities promising results achieved neighborhood convolutional network model &# 8217 limited processing capabilities depthwise separable convolution high segmentation accuracy 3d organ segmentation routine clinical workflows memory demands make net builds upon entirely new architecture comparative analysis verified comparative analysis manual segmentation memory usage clinical deployment viable alternative time applications recent years purpose due ongoing efforts net variant net framework less suitable inference speed including upernet highly suitable first time findings contribute explore post edge devices constrained environments computed tomography 2d convolutions |