An Improved 3D Deep Learning-Based Segmentation of Left Ventricular Myocardial Diseases from Delayed-Enhancement MRI with Inclusion and Classification Prior Information U-Net (ICPIU-Net)
Accurate segmentation of the myocardial scar may supply relevant advancements in predicting and controlling deadly ventricular arrhythmias in subjects with cardiovascular disease. In this paper, we propose the architecture of inclusion and classification of prior information U-Net (ICPIU-Net) to eff...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-03-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/6/2084 |
| _version_ | 1797442505791242240 |
|---|---|
| author | Khawla Brahim Tewodros Weldebirhan Arega Arnaud Boucher Stephanie Bricq Anis Sakly Fabrice Meriaudeau |
| author_facet | Khawla Brahim Tewodros Weldebirhan Arega Arnaud Boucher Stephanie Bricq Anis Sakly Fabrice Meriaudeau |
| author_sort | Khawla Brahim |
| collection | DOAJ |
| description | Accurate segmentation of the myocardial scar may supply relevant advancements in predicting and controlling deadly ventricular arrhythmias in subjects with cardiovascular disease. In this paper, we propose the architecture of inclusion and classification of prior information U-Net (ICPIU-Net) to efficiently segment the left ventricle (LV) myocardium, myocardial infarction (MI), and microvascular-obstructed (MVO) tissues from late gadolinium enhancement magnetic resonance (LGE-MR) images. Our approach was developed using two subnets cascaded to first segment the LV cavity and myocardium. Then, we used inclusion and classification constraint networks to improve the resulting segmentation of the diseased regions within the pre-segmented LV myocardium. This network incorporates the inclusion and classification information of the LGE-MRI to maintain topological constraints of pathological areas. In the testing stage, the outputs of each segmentation network obtained with specific estimated parameters from training were fused using the majority voting technique for the final label prediction of each voxel in the LGE-MR image. The proposed method was validated by comparing its results to manual drawings by experts from 50 LGE-MR images. Importantly, compared to various deep learning-based methods participating in the EMIDEC challenge, the results of our approach have a more significant agreement with manual contouring in segmenting myocardial diseases. |
| first_indexed | 2024-03-09T12:42:50Z |
| format | Article |
| id | doaj.art-6a3608e2a382443ba1fd139c5a11004c |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T12:42:50Z |
| publishDate | 2022-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-6a3608e2a382443ba1fd139c5a11004c2023-11-30T22:16:00ZengMDPI AGSensors1424-82202022-03-01226208410.3390/s22062084An Improved 3D Deep Learning-Based Segmentation of Left Ventricular Myocardial Diseases from Delayed-Enhancement MRI with Inclusion and Classification Prior Information U-Net (ICPIU-Net)Khawla Brahim0Tewodros Weldebirhan Arega1Arnaud Boucher2Stephanie Bricq3Anis Sakly4Fabrice Meriaudeau5ImViA EA 7535 Laboratory, University of Burgundy, 21078 Dijon, FranceImViA EA 7535 Laboratory, University of Burgundy, 21078 Dijon, FranceImViA EA 7535 Laboratory, University of Burgundy, 21078 Dijon, FranceImViA EA 7535 Laboratory, University of Burgundy, 21078 Dijon, FranceLASEE Laboratory, National Engineering School of Monastir, University of Monastir, Monastir 5000, TunisiaImViA EA 7535 Laboratory, University of Burgundy, 21078 Dijon, FranceAccurate segmentation of the myocardial scar may supply relevant advancements in predicting and controlling deadly ventricular arrhythmias in subjects with cardiovascular disease. In this paper, we propose the architecture of inclusion and classification of prior information U-Net (ICPIU-Net) to efficiently segment the left ventricle (LV) myocardium, myocardial infarction (MI), and microvascular-obstructed (MVO) tissues from late gadolinium enhancement magnetic resonance (LGE-MR) images. Our approach was developed using two subnets cascaded to first segment the LV cavity and myocardium. Then, we used inclusion and classification constraint networks to improve the resulting segmentation of the diseased regions within the pre-segmented LV myocardium. This network incorporates the inclusion and classification information of the LGE-MRI to maintain topological constraints of pathological areas. In the testing stage, the outputs of each segmentation network obtained with specific estimated parameters from training were fused using the majority voting technique for the final label prediction of each voxel in the LGE-MR image. The proposed method was validated by comparing its results to manual drawings by experts from 50 LGE-MR images. Importantly, compared to various deep learning-based methods participating in the EMIDEC challenge, the results of our approach have a more significant agreement with manual contouring in segmenting myocardial diseases.https://www.mdpi.com/1424-8220/22/6/2084segmentationICPIU-Netmyocardiummyocardial infarction (MI)late gadolinium enhancement magnetic resonance (LGE-MR)microvascular-obstructed (MVO) |
| spellingShingle | Khawla Brahim Tewodros Weldebirhan Arega Arnaud Boucher Stephanie Bricq Anis Sakly Fabrice Meriaudeau An Improved 3D Deep Learning-Based Segmentation of Left Ventricular Myocardial Diseases from Delayed-Enhancement MRI with Inclusion and Classification Prior Information U-Net (ICPIU-Net) Sensors segmentation ICPIU-Net myocardium myocardial infarction (MI) late gadolinium enhancement magnetic resonance (LGE-MR) microvascular-obstructed (MVO) |
| title | An Improved 3D Deep Learning-Based Segmentation of Left Ventricular Myocardial Diseases from Delayed-Enhancement MRI with Inclusion and Classification Prior Information U-Net (ICPIU-Net) |
| title_full | An Improved 3D Deep Learning-Based Segmentation of Left Ventricular Myocardial Diseases from Delayed-Enhancement MRI with Inclusion and Classification Prior Information U-Net (ICPIU-Net) |
| title_fullStr | An Improved 3D Deep Learning-Based Segmentation of Left Ventricular Myocardial Diseases from Delayed-Enhancement MRI with Inclusion and Classification Prior Information U-Net (ICPIU-Net) |
| title_full_unstemmed | An Improved 3D Deep Learning-Based Segmentation of Left Ventricular Myocardial Diseases from Delayed-Enhancement MRI with Inclusion and Classification Prior Information U-Net (ICPIU-Net) |
| title_short | An Improved 3D Deep Learning-Based Segmentation of Left Ventricular Myocardial Diseases from Delayed-Enhancement MRI with Inclusion and Classification Prior Information U-Net (ICPIU-Net) |
| title_sort | improved 3d deep learning based segmentation of left ventricular myocardial diseases from delayed enhancement mri with inclusion and classification prior information u net icpiu net |
| topic | segmentation ICPIU-Net myocardium myocardial infarction (MI) late gadolinium enhancement magnetic resonance (LGE-MR) microvascular-obstructed (MVO) |
| url | https://www.mdpi.com/1424-8220/22/6/2084 |
| work_keys_str_mv | AT khawlabrahim animproved3ddeeplearningbasedsegmentationofleftventricularmyocardialdiseasesfromdelayedenhancementmriwithinclusionandclassificationpriorinformationuneticpiunet AT tewodrosweldebirhanarega animproved3ddeeplearningbasedsegmentationofleftventricularmyocardialdiseasesfromdelayedenhancementmriwithinclusionandclassificationpriorinformationuneticpiunet AT arnaudboucher animproved3ddeeplearningbasedsegmentationofleftventricularmyocardialdiseasesfromdelayedenhancementmriwithinclusionandclassificationpriorinformationuneticpiunet AT stephaniebricq animproved3ddeeplearningbasedsegmentationofleftventricularmyocardialdiseasesfromdelayedenhancementmriwithinclusionandclassificationpriorinformationuneticpiunet AT anissakly animproved3ddeeplearningbasedsegmentationofleftventricularmyocardialdiseasesfromdelayedenhancementmriwithinclusionandclassificationpriorinformationuneticpiunet AT fabricemeriaudeau animproved3ddeeplearningbasedsegmentationofleftventricularmyocardialdiseasesfromdelayedenhancementmriwithinclusionandclassificationpriorinformationuneticpiunet AT khawlabrahim improved3ddeeplearningbasedsegmentationofleftventricularmyocardialdiseasesfromdelayedenhancementmriwithinclusionandclassificationpriorinformationuneticpiunet AT tewodrosweldebirhanarega improved3ddeeplearningbasedsegmentationofleftventricularmyocardialdiseasesfromdelayedenhancementmriwithinclusionandclassificationpriorinformationuneticpiunet AT arnaudboucher improved3ddeeplearningbasedsegmentationofleftventricularmyocardialdiseasesfromdelayedenhancementmriwithinclusionandclassificationpriorinformationuneticpiunet AT stephaniebricq improved3ddeeplearningbasedsegmentationofleftventricularmyocardialdiseasesfromdelayedenhancementmriwithinclusionandclassificationpriorinformationuneticpiunet AT anissakly improved3ddeeplearningbasedsegmentationofleftventricularmyocardialdiseasesfromdelayedenhancementmriwithinclusionandclassificationpriorinformationuneticpiunet AT fabricemeriaudeau improved3ddeeplearningbasedsegmentationofleftventricularmyocardialdiseasesfromdelayedenhancementmriwithinclusionandclassificationpriorinformationuneticpiunet |