Automatic generation of subject-specific finite element models of the spine from magnetic resonance images
The generation of subject-specific finite element models of the spine is generally a time-consuming process based on computed tomography (CT) images, where scanning exposes subjects to harmful radiation. In this study, a method is presented for the automatic generation of spine finite element models...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2023-09-01
|
| Series: | Frontiers in Bioengineering and Biotechnology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fbioe.2023.1244291/full |
| _version_ | 1797692541866344448 |
|---|---|
| author | Joeri Kok Yulia M. Shcherbakova Tom P. C. Schlösser Peter R. Seevinck Peter R. Seevinck Tijl A. van der Velden Tijl A. van der Velden René M. Castelein Keita Ito Keita Ito Bert van Rietbergen |
| author_facet | Joeri Kok Yulia M. Shcherbakova Tom P. C. Schlösser Peter R. Seevinck Peter R. Seevinck Tijl A. van der Velden Tijl A. van der Velden René M. Castelein Keita Ito Keita Ito Bert van Rietbergen |
| author_sort | Joeri Kok |
| collection | DOAJ |
| description | The generation of subject-specific finite element models of the spine is generally a time-consuming process based on computed tomography (CT) images, where scanning exposes subjects to harmful radiation. In this study, a method is presented for the automatic generation of spine finite element models using images from a single magnetic resonance (MR) sequence. The thoracic and lumbar spine of eight adult volunteers was imaged using a 3D multi-echo-gradient-echo sagittal MR sequence. A deep-learning method was used to generate synthetic CT images from the MR images. A pre-trained deep-learning network was used for the automatic segmentation of vertebrae from the synthetic CT images. Another deep-learning network was trained for the automatic segmentation of intervertebral discs from the MR images. The automatic segmentations were validated against manual segmentations for two subjects, one with scoliosis, and another with a spine implant. A template mesh of the spine was registered to the segmentations in three steps using a Bayesian coherent point drift algorithm. First, rigid registration was applied on the complete spine. Second, non-rigid registration was used for the individual discs and vertebrae. Third, the complete spine was non-rigidly registered to the individually registered discs and vertebrae. Comparison of the automatic and manual segmentations led to dice-scores of 0.93–0.96 for all vertebrae and discs. The lowest dice-score was in the disc at the height of the implant where artifacts led to under-segmentation. The mean distance between the morphed meshes and the segmentations was below 1 mm. In conclusion, the presented method can be used to automatically generate accurate subject-specific spine models. |
| first_indexed | 2024-03-12T02:29:09Z |
| format | Article |
| id | doaj.art-8f26a1f66c0e4e6989f4bf5ef5848ea7 |
| institution | Directory Open Access Journal |
| issn | 2296-4185 |
| language | English |
| last_indexed | 2024-03-12T02:29:09Z |
| publishDate | 2023-09-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Bioengineering and Biotechnology |
| spelling | doaj.art-8f26a1f66c0e4e6989f4bf5ef5848ea72023-09-05T09:20:47ZengFrontiers Media S.A.Frontiers in Bioengineering and Biotechnology2296-41852023-09-011110.3389/fbioe.2023.12442911244291Automatic generation of subject-specific finite element models of the spine from magnetic resonance imagesJoeri Kok0Yulia M. Shcherbakova1Tom P. C. Schlösser2Peter R. Seevinck3Peter R. Seevinck4Tijl A. van der Velden5Tijl A. van der Velden6René M. Castelein7Keita Ito8Keita Ito9Bert van Rietbergen10Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, NetherlandsImage Sciences Institute, University Medical Center Utrecht, Utrecht, NetherlandsDepartment of Orthopaedic Surgery, University Medical Center Utrecht, Utrecht, NetherlandsImage Sciences Institute, University Medical Center Utrecht, Utrecht, NetherlandsMRIguidance BV, Utrecht, NetherlandsImage Sciences Institute, University Medical Center Utrecht, Utrecht, NetherlandsMRIguidance BV, Utrecht, NetherlandsDepartment of Orthopaedic Surgery, University Medical Center Utrecht, Utrecht, NetherlandsDepartment of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, NetherlandsDepartment of Orthopaedic Surgery, University Medical Center Utrecht, Utrecht, NetherlandsDepartment of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, NetherlandsThe generation of subject-specific finite element models of the spine is generally a time-consuming process based on computed tomography (CT) images, where scanning exposes subjects to harmful radiation. In this study, a method is presented for the automatic generation of spine finite element models using images from a single magnetic resonance (MR) sequence. The thoracic and lumbar spine of eight adult volunteers was imaged using a 3D multi-echo-gradient-echo sagittal MR sequence. A deep-learning method was used to generate synthetic CT images from the MR images. A pre-trained deep-learning network was used for the automatic segmentation of vertebrae from the synthetic CT images. Another deep-learning network was trained for the automatic segmentation of intervertebral discs from the MR images. The automatic segmentations were validated against manual segmentations for two subjects, one with scoliosis, and another with a spine implant. A template mesh of the spine was registered to the segmentations in three steps using a Bayesian coherent point drift algorithm. First, rigid registration was applied on the complete spine. Second, non-rigid registration was used for the individual discs and vertebrae. Third, the complete spine was non-rigidly registered to the individually registered discs and vertebrae. Comparison of the automatic and manual segmentations led to dice-scores of 0.93–0.96 for all vertebrae and discs. The lowest dice-score was in the disc at the height of the implant where artifacts led to under-segmentation. The mean distance between the morphed meshes and the segmentations was below 1 mm. In conclusion, the presented method can be used to automatically generate accurate subject-specific spine models.https://www.frontiersin.org/articles/10.3389/fbioe.2023.1244291/fullsynthetic computed tomographydeep-learningmesh morphingpersonalized medicinevertebraintervertebral disc |
| spellingShingle | Joeri Kok Yulia M. Shcherbakova Tom P. C. Schlösser Peter R. Seevinck Peter R. Seevinck Tijl A. van der Velden Tijl A. van der Velden René M. Castelein Keita Ito Keita Ito Bert van Rietbergen Automatic generation of subject-specific finite element models of the spine from magnetic resonance images Frontiers in Bioengineering and Biotechnology synthetic computed tomography deep-learning mesh morphing personalized medicine vertebra intervertebral disc |
| title | Automatic generation of subject-specific finite element models of the spine from magnetic resonance images |
| title_full | Automatic generation of subject-specific finite element models of the spine from magnetic resonance images |
| title_fullStr | Automatic generation of subject-specific finite element models of the spine from magnetic resonance images |
| title_full_unstemmed | Automatic generation of subject-specific finite element models of the spine from magnetic resonance images |
| title_short | Automatic generation of subject-specific finite element models of the spine from magnetic resonance images |
| title_sort | automatic generation of subject specific finite element models of the spine from magnetic resonance images |
| topic | synthetic computed tomography deep-learning mesh morphing personalized medicine vertebra intervertebral disc |
| url | https://www.frontiersin.org/articles/10.3389/fbioe.2023.1244291/full |
| work_keys_str_mv | AT joerikok automaticgenerationofsubjectspecificfiniteelementmodelsofthespinefrommagneticresonanceimages AT yuliamshcherbakova automaticgenerationofsubjectspecificfiniteelementmodelsofthespinefrommagneticresonanceimages AT tompcschlosser automaticgenerationofsubjectspecificfiniteelementmodelsofthespinefrommagneticresonanceimages AT peterrseevinck automaticgenerationofsubjectspecificfiniteelementmodelsofthespinefrommagneticresonanceimages AT peterrseevinck automaticgenerationofsubjectspecificfiniteelementmodelsofthespinefrommagneticresonanceimages AT tijlavandervelden automaticgenerationofsubjectspecificfiniteelementmodelsofthespinefrommagneticresonanceimages AT tijlavandervelden automaticgenerationofsubjectspecificfiniteelementmodelsofthespinefrommagneticresonanceimages AT renemcastelein automaticgenerationofsubjectspecificfiniteelementmodelsofthespinefrommagneticresonanceimages AT keitaito automaticgenerationofsubjectspecificfiniteelementmodelsofthespinefrommagneticresonanceimages AT keitaito automaticgenerationofsubjectspecificfiniteelementmodelsofthespinefrommagneticresonanceimages AT bertvanrietbergen automaticgenerationofsubjectspecificfiniteelementmodelsofthespinefrommagneticresonanceimages |