SIP-UNet: Sequential Inputs Parallel UNet Architecture for Segmentation of Brain Tissues from Magnetic Resonance Images
Proper analysis of changes in brain structure can lead to a more accurate diagnosis of specific brain disorders. The accuracy of segmentation is crucial for quantifying changes in brain structure. In recent studies, UNet-based architectures have outperformed other deep learning architectures in biom...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-08-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/10/15/2755 |
| _version_ | 1797441413191827456 |
|---|---|
| author | Rukesh Prajapati Goo-Rak Kwon |
| author_facet | Rukesh Prajapati Goo-Rak Kwon |
| author_sort | Rukesh Prajapati |
| collection | DOAJ |
| description | Proper analysis of changes in brain structure can lead to a more accurate diagnosis of specific brain disorders. The accuracy of segmentation is crucial for quantifying changes in brain structure. In recent studies, UNet-based architectures have outperformed other deep learning architectures in biomedical image segmentation. However, improving segmentation accuracy is challenging due to the low resolution of medical images and insufficient data. In this study, we present a novel architecture that combines three parallel UNets using a residual network. This architecture improves upon the baseline methods in three ways. First, instead of using a single image as input, we use three consecutive images. This gives our model the freedom to learn from neighboring images as well. Additionally, the images are individually compressed and decompressed using three different UNets, which prevents the model from merging the features of the images. Finally, following the residual network architecture, the outputs of the UNets are combined in such a way that the features of the image corresponding to the output are enhanced by a skip connection. The proposed architecture performed better than using a single conventional UNet and other UNet variants. |
| first_indexed | 2024-03-09T12:22:42Z |
| format | Article |
| id | doaj.art-50c5828900314f7285bea299bc7d867b |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-09T12:22:42Z |
| publishDate | 2022-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-50c5828900314f7285bea299bc7d867b2023-11-30T22:38:44ZengMDPI AGMathematics2227-73902022-08-011015275510.3390/math10152755SIP-UNet: Sequential Inputs Parallel UNet Architecture for Segmentation of Brain Tissues from Magnetic Resonance ImagesRukesh Prajapati0Goo-Rak Kwon1Department of Information and Communication Engineering, Chosun University, 309 Pilmun-Daero, Dong-Gu, Gwangju 61452, KoreaDepartment of Information and Communication Engineering, Chosun University, 309 Pilmun-Daero, Dong-Gu, Gwangju 61452, KoreaProper analysis of changes in brain structure can lead to a more accurate diagnosis of specific brain disorders. The accuracy of segmentation is crucial for quantifying changes in brain structure. In recent studies, UNet-based architectures have outperformed other deep learning architectures in biomedical image segmentation. However, improving segmentation accuracy is challenging due to the low resolution of medical images and insufficient data. In this study, we present a novel architecture that combines three parallel UNets using a residual network. This architecture improves upon the baseline methods in three ways. First, instead of using a single image as input, we use three consecutive images. This gives our model the freedom to learn from neighboring images as well. Additionally, the images are individually compressed and decompressed using three different UNets, which prevents the model from merging the features of the images. Finally, following the residual network architecture, the outputs of the UNets are combined in such a way that the features of the image corresponding to the output are enhanced by a skip connection. The proposed architecture performed better than using a single conventional UNet and other UNet variants.https://www.mdpi.com/2227-7390/10/15/2755biomedical image segmentationdeep learningparallel UNetResNet |
| spellingShingle | Rukesh Prajapati Goo-Rak Kwon SIP-UNet: Sequential Inputs Parallel UNet Architecture for Segmentation of Brain Tissues from Magnetic Resonance Images Mathematics biomedical image segmentation deep learning parallel UNet ResNet |
| title | SIP-UNet: Sequential Inputs Parallel UNet Architecture for Segmentation of Brain Tissues from Magnetic Resonance Images |
| title_full | SIP-UNet: Sequential Inputs Parallel UNet Architecture for Segmentation of Brain Tissues from Magnetic Resonance Images |
| title_fullStr | SIP-UNet: Sequential Inputs Parallel UNet Architecture for Segmentation of Brain Tissues from Magnetic Resonance Images |
| title_full_unstemmed | SIP-UNet: Sequential Inputs Parallel UNet Architecture for Segmentation of Brain Tissues from Magnetic Resonance Images |
| title_short | SIP-UNet: Sequential Inputs Parallel UNet Architecture for Segmentation of Brain Tissues from Magnetic Resonance Images |
| title_sort | sip unet sequential inputs parallel unet architecture for segmentation of brain tissues from magnetic resonance images |
| topic | biomedical image segmentation deep learning parallel UNet ResNet |
| url | https://www.mdpi.com/2227-7390/10/15/2755 |
| work_keys_str_mv | AT rukeshprajapati sipunetsequentialinputsparallelunetarchitectureforsegmentationofbraintissuesfrommagneticresonanceimages AT goorakkwon sipunetsequentialinputsparallelunetarchitectureforsegmentationofbraintissuesfrommagneticresonanceimages |