Analysis of U-Net Neural Network Training Parameters for Tomographic Images Segmentation
Image segmentation is one of the main resources in computer vision. Nowadays, this procedure can be made with high precision using Deep Learning, and this fact is important to applications of several research areas including medical image analysis. Image segmentation is currently applied to find tum...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Escuela Politécnica Nacional (EPN)
2023-07-01
|
| Series: | Latin-American Journal of Computing |
| Subjects: | |
| Online Access: | https://lajc.epn.edu.ec/index.php/LAJC/article/view/359 |
| _version_ | 1797692267792695296 |
|---|---|
| author | Yana dos Santos Pereira Davi Guimarães da Silva Regina Cely Barroso Anderson Alvarenga de Moura Meneses |
| author_facet | Yana dos Santos Pereira Davi Guimarães da Silva Regina Cely Barroso Anderson Alvarenga de Moura Meneses |
| author_sort | Yana dos Santos Pereira |
| collection | DOAJ |
| description | Image segmentation is one of the main resources in computer vision. Nowadays, this procedure can be made with high precision using Deep Learning, and this fact is important to applications of several research areas including medical image analysis. Image segmentation is currently applied to find tumors, bone defects and other elements that are crucial to achieve accurate diagnoses. The objective of the present work is to verify the influence of parameters variation on U-Net, a Deep Convolutional Neural Network with Deep Learning for biomedical image segmentation. The dataset was obtained from Kaggle website (www.kaggle.com) and contains 267 volumes of lung computed tomography scans, which are composed of the 2D images and their respective masks (ground truth). The dataset was subdivided in 80% of the volumes for training and 20% for testing. The results were evaluated using the Dice Similarity Coefficient as metric and the value 84% was the mean obtained for the testing set, applying the best parameters considered. |
| first_indexed | 2024-03-12T02:25:11Z |
| format | Article |
| id | doaj.art-87f0c61a728f4624b48153375e8ec7cb |
| institution | Directory Open Access Journal |
| issn | 1390-9266 1390-9134 |
| language | English |
| last_indexed | 2024-03-12T02:25:11Z |
| publishDate | 2023-07-01 |
| publisher | Escuela Politécnica Nacional (EPN) |
| record_format | Article |
| series | Latin-American Journal of Computing |
| spelling | doaj.art-87f0c61a728f4624b48153375e8ec7cb2023-09-05T14:48:20ZengEscuela Politécnica Nacional (EPN)Latin-American Journal of Computing1390-92661390-91342023-07-011028495359Analysis of U-Net Neural Network Training Parameters for Tomographic Images SegmentationYana dos Santos Pereira0Davi Guimarães da Silva1Regina Cely Barroso2Anderson Alvarenga de Moura Meneses3Federal University of Western ParáFederal University of Western ParáUniversity of Rio de Janeiro Rio de Janeiro, BrazilFederal University of Western ParáImage segmentation is one of the main resources in computer vision. Nowadays, this procedure can be made with high precision using Deep Learning, and this fact is important to applications of several research areas including medical image analysis. Image segmentation is currently applied to find tumors, bone defects and other elements that are crucial to achieve accurate diagnoses. The objective of the present work is to verify the influence of parameters variation on U-Net, a Deep Convolutional Neural Network with Deep Learning for biomedical image segmentation. The dataset was obtained from Kaggle website (www.kaggle.com) and contains 267 volumes of lung computed tomography scans, which are composed of the 2D images and their respective masks (ground truth). The dataset was subdivided in 80% of the volumes for training and 20% for testing. The results were evaluated using the Dice Similarity Coefficient as metric and the value 84% was the mean obtained for the testing set, applying the best parameters considered.https://lajc.epn.edu.ec/index.php/LAJC/article/view/359deep learningbiomedical image segmentationfully convolutional networksu-netcomputed tomography |
| spellingShingle | Yana dos Santos Pereira Davi Guimarães da Silva Regina Cely Barroso Anderson Alvarenga de Moura Meneses Analysis of U-Net Neural Network Training Parameters for Tomographic Images Segmentation Latin-American Journal of Computing deep learning biomedical image segmentation fully convolutional networks u-net computed tomography |
| title | Analysis of U-Net Neural Network Training Parameters for Tomographic Images Segmentation |
| title_full | Analysis of U-Net Neural Network Training Parameters for Tomographic Images Segmentation |
| title_fullStr | Analysis of U-Net Neural Network Training Parameters for Tomographic Images Segmentation |
| title_full_unstemmed | Analysis of U-Net Neural Network Training Parameters for Tomographic Images Segmentation |
| title_short | Analysis of U-Net Neural Network Training Parameters for Tomographic Images Segmentation |
| title_sort | analysis of u net neural network training parameters for tomographic images segmentation |
| topic | deep learning biomedical image segmentation fully convolutional networks u-net computed tomography |
| url | https://lajc.epn.edu.ec/index.php/LAJC/article/view/359 |
| work_keys_str_mv | AT yanadossantospereira analysisofunetneuralnetworktrainingparametersfortomographicimagessegmentation AT daviguimaraesdasilva analysisofunetneuralnetworktrainingparametersfortomographicimagessegmentation AT reginacelybarroso analysisofunetneuralnetworktrainingparametersfortomographicimagessegmentation AT andersonalvarengademourameneses analysisofunetneuralnetworktrainingparametersfortomographicimagessegmentation |