Deep Learning Based on EfficientNet for Multiorgan Segmentation of Thoracic Structures on a 0.35 T MR-Linac Radiation Therapy System
The advent of the 0.35 T MR-Linac (MRIdian, ViewRay) system in radiation therapy allows precise tumor targeting for moving lesions. However, the lack of an automatic volume segmentation function in the MR-Linac’s treatment planning system poses a challenge. In this paper, we propose a deep-learning-...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-12-01
|
Series: | Algorithms |
Subjects: | |
Online Access: | https://www.mdpi.com/1999-4893/16/12/564 |
_version_ | 1797382284153716736 |
---|---|
author | Mohammed Chekroun Youssef Mourchid Igor Bessières Alain Lalande |
author_facet | Mohammed Chekroun Youssef Mourchid Igor Bessières Alain Lalande |
author_sort | Mohammed Chekroun |
collection | DOAJ |
description | The advent of the 0.35 T MR-Linac (MRIdian, ViewRay) system in radiation therapy allows precise tumor targeting for moving lesions. However, the lack of an automatic volume segmentation function in the MR-Linac’s treatment planning system poses a challenge. In this paper, we propose a deep-learning-based multiorgan segmentation approach for the thoracic region, using EfficientNet as the backbone for the network architecture. The objectives of this approach include accurate segmentation of critical organs, such as the left and right lungs, the heart, the spinal cord, and the esophagus, essential for minimizing radiation toxicity during external radiation therapy. Our proposed approach, when evaluated on an internal dataset comprising 81 patients, demonstrated superior performance compared to other state-of-the-art methods. Specifically, the results for our approach with a 2.5D strategy were as follows: a dice similarity coefficient (DSC) of 0.820 ± 0.041, an intersection over union (IoU) of 0.725 ± 0.052, and a 3D Hausdorff distance (HD) of 10.353 ± 4.974 mm. Notably, the 2.5D strategy surpassed the 2D strategy in all three metrics, exhibiting higher DSC and IoU values, as well as lower HD values. This improvement strongly suggests that our proposed approach with the 2.5D strategy may hold promise in achieving more precise and accurate segmentations when compared to the conventional 2D strategy. Our work has practical implications in the improvement of treatment planning precision, aligning with the evolution of medical imaging and innovative strategies for multiorgan segmentation tasks. |
first_indexed | 2024-03-08T21:03:13Z |
format | Article |
id | doaj.art-16070e6d1c8d421cbc5aa228ebab0dcf |
institution | Directory Open Access Journal |
issn | 1999-4893 |
language | English |
last_indexed | 2024-03-08T21:03:13Z |
publishDate | 2023-12-01 |
publisher | MDPI AG |
record_format | Article |
series | Algorithms |
spelling | doaj.art-16070e6d1c8d421cbc5aa228ebab0dcf2023-12-22T13:47:05ZengMDPI AGAlgorithms1999-48932023-12-01161256410.3390/a16120564Deep Learning Based on EfficientNet for Multiorgan Segmentation of Thoracic Structures on a 0.35 T MR-Linac Radiation Therapy SystemMohammed Chekroun0Youssef Mourchid1Igor Bessières2Alain Lalande3CESI LINEACT Laboratory, UR 7527, 21000 Dijon, FranceCESI LINEACT Laboratory, UR 7527, 21000 Dijon, FranceCentre Georges-François Leclerc, 21000 Dijon, FranceIFTIM, ICMUB Laboratory, UMR CNRS 6302, University of Burgundy, 21000 Dijon, FranceThe advent of the 0.35 T MR-Linac (MRIdian, ViewRay) system in radiation therapy allows precise tumor targeting for moving lesions. However, the lack of an automatic volume segmentation function in the MR-Linac’s treatment planning system poses a challenge. In this paper, we propose a deep-learning-based multiorgan segmentation approach for the thoracic region, using EfficientNet as the backbone for the network architecture. The objectives of this approach include accurate segmentation of critical organs, such as the left and right lungs, the heart, the spinal cord, and the esophagus, essential for minimizing radiation toxicity during external radiation therapy. Our proposed approach, when evaluated on an internal dataset comprising 81 patients, demonstrated superior performance compared to other state-of-the-art methods. Specifically, the results for our approach with a 2.5D strategy were as follows: a dice similarity coefficient (DSC) of 0.820 ± 0.041, an intersection over union (IoU) of 0.725 ± 0.052, and a 3D Hausdorff distance (HD) of 10.353 ± 4.974 mm. Notably, the 2.5D strategy surpassed the 2D strategy in all three metrics, exhibiting higher DSC and IoU values, as well as lower HD values. This improvement strongly suggests that our proposed approach with the 2.5D strategy may hold promise in achieving more precise and accurate segmentations when compared to the conventional 2D strategy. Our work has practical implications in the improvement of treatment planning precision, aligning with the evolution of medical imaging and innovative strategies for multiorgan segmentation tasks.https://www.mdpi.com/1999-4893/16/12/564organs-at-risksegmentationMRIdeep learningradiation therapy0.35 T MR-Linac |
spellingShingle | Mohammed Chekroun Youssef Mourchid Igor Bessières Alain Lalande Deep Learning Based on EfficientNet for Multiorgan Segmentation of Thoracic Structures on a 0.35 T MR-Linac Radiation Therapy System Algorithms organs-at-risk segmentation MRI deep learning radiation therapy 0.35 T MR-Linac |
title | Deep Learning Based on EfficientNet for Multiorgan Segmentation of Thoracic Structures on a 0.35 T MR-Linac Radiation Therapy System |
title_full | Deep Learning Based on EfficientNet for Multiorgan Segmentation of Thoracic Structures on a 0.35 T MR-Linac Radiation Therapy System |
title_fullStr | Deep Learning Based on EfficientNet for Multiorgan Segmentation of Thoracic Structures on a 0.35 T MR-Linac Radiation Therapy System |
title_full_unstemmed | Deep Learning Based on EfficientNet for Multiorgan Segmentation of Thoracic Structures on a 0.35 T MR-Linac Radiation Therapy System |
title_short | Deep Learning Based on EfficientNet for Multiorgan Segmentation of Thoracic Structures on a 0.35 T MR-Linac Radiation Therapy System |
title_sort | deep learning based on efficientnet for multiorgan segmentation of thoracic structures on a 0 35 t mr linac radiation therapy system |
topic | organs-at-risk segmentation MRI deep learning radiation therapy 0.35 T MR-Linac |
url | https://www.mdpi.com/1999-4893/16/12/564 |
work_keys_str_mv | AT mohammedchekroun deeplearningbasedonefficientnetformultiorgansegmentationofthoracicstructuresona035tmrlinacradiationtherapysystem AT youssefmourchid deeplearningbasedonefficientnetformultiorgansegmentationofthoracicstructuresona035tmrlinacradiationtherapysystem AT igorbessieres deeplearningbasedonefficientnetformultiorgansegmentationofthoracicstructuresona035tmrlinacradiationtherapysystem AT alainlalande deeplearningbasedonefficientnetformultiorgansegmentationofthoracicstructuresona035tmrlinacradiationtherapysystem |