Diagnosis of osteoporotic vertebral compression fractures and fracture level detection using multitask learning with U-Net in lumbar spine lateral radiographs
Recent studies of automatic diagnosis of vertebral compression fractures (VCFs) using deep learning mainly focus on segmentation and vertebral level detection in lumbar spine lateral radiographs (LSLRs). Herein, we developed a model for simultaneous VCF diagnosis and vertebral level detection withou...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2023-01-01
|
| Series: | Computational and Structural Biotechnology Journal |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2001037023002301 |
| _version_ | 1797384098676736000 |
|---|---|
| author | Seung Min Ryu Soyoung Lee Miso Jang Jung-Min Koh Sung Jin Bae Seong Gyu Jegal Keewon Shin Namkug Kim |
| author_facet | Seung Min Ryu Soyoung Lee Miso Jang Jung-Min Koh Sung Jin Bae Seong Gyu Jegal Keewon Shin Namkug Kim |
| author_sort | Seung Min Ryu |
| collection | DOAJ |
| description | Recent studies of automatic diagnosis of vertebral compression fractures (VCFs) using deep learning mainly focus on segmentation and vertebral level detection in lumbar spine lateral radiographs (LSLRs). Herein, we developed a model for simultaneous VCF diagnosis and vertebral level detection without using adjacent vertebral bodies. In total, 1102 patients with VCF, 1171 controls were enrolled. The 1865, 208, and 198 LSLRS were divided into training, validation, and test dataset. A ground truth label with a 4-point trapezoidal shape was made based on radiological reports showing normal or VCF at some vertebral level. We applied a modified U-Net architecture, in which decoders were trained to detect VCF and vertebral levels, sharing the same encoder. The multi-task model was significantly better than the single-task model in sensitivity and area under the receiver operating characteristic curve. In the internal dataset, the accuracy, sensitivity, and specificity of fracture detection per patient or vertebral body were 0.929, 0.944, and 0.917 or 0.947, 0.628, and 0.977, respectively. In external validation, those of fracture detection per patient or vertebral body were 0.713, 0.979, and 0.447 or 0.828, 0.936, and 0.820, respectively. The success rates were 96 % and 94 % for vertebral level detection in internal and external validation, respectively. The multi-task-shared encoder was significantly better than the single-task encoder. Furthermore, both fracture and vertebral level detection was good in internal and external validation. Our deep learning model may help radiologists perform real-life medical examinations. |
| first_indexed | 2024-03-08T21:30:35Z |
| format | Article |
| id | doaj.art-d78bb8ed18144f85b6851f26918f07ea |
| institution | Directory Open Access Journal |
| issn | 2001-0370 |
| language | English |
| last_indexed | 2024-03-08T21:30:35Z |
| publishDate | 2023-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Computational and Structural Biotechnology Journal |
| spelling | doaj.art-d78bb8ed18144f85b6851f26918f07ea2023-12-21T07:31:42ZengElsevierComputational and Structural Biotechnology Journal2001-03702023-01-012134523458Diagnosis of osteoporotic vertebral compression fractures and fracture level detection using multitask learning with U-Net in lumbar spine lateral radiographsSeung Min Ryu0Soyoung Lee1Miso Jang2Jung-Min Koh3Sung Jin Bae4Seong Gyu Jegal5Keewon Shin6Namkug Kim7Department of Orthopedic Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea; Department of Biomedical Engineering, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of KoreaDepartment of Biomedical Engineering, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of KoreaDepartment of Convergence Medicine, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of KoreaDivision of Endocrinology and Metabolism, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of KoreaDepartment of Health Screening and Promotion Center, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of KoreaDepartment of Biomedical Engineering, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of KoreaDepartment of Biomedical Engineering, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea; Correspondence to: Department of Biomedical Engineering, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, 4F, 26, Olympic-ro 43-gil, Songpa-gu, Seoul 05505, Republic of Korea.Department of Biomedical Engineering, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea; Department of Convergence Medicine, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea; Correspondence to: Department of Convergence Medicine, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, 5F, 26, Olympic-ro 43-gil, Songpa-gu, Seoul 05505, Republic of Korea.Recent studies of automatic diagnosis of vertebral compression fractures (VCFs) using deep learning mainly focus on segmentation and vertebral level detection in lumbar spine lateral radiographs (LSLRs). Herein, we developed a model for simultaneous VCF diagnosis and vertebral level detection without using adjacent vertebral bodies. In total, 1102 patients with VCF, 1171 controls were enrolled. The 1865, 208, and 198 LSLRS were divided into training, validation, and test dataset. A ground truth label with a 4-point trapezoidal shape was made based on radiological reports showing normal or VCF at some vertebral level. We applied a modified U-Net architecture, in which decoders were trained to detect VCF and vertebral levels, sharing the same encoder. The multi-task model was significantly better than the single-task model in sensitivity and area under the receiver operating characteristic curve. In the internal dataset, the accuracy, sensitivity, and specificity of fracture detection per patient or vertebral body were 0.929, 0.944, and 0.917 or 0.947, 0.628, and 0.977, respectively. In external validation, those of fracture detection per patient or vertebral body were 0.713, 0.979, and 0.447 or 0.828, 0.936, and 0.820, respectively. The success rates were 96 % and 94 % for vertebral level detection in internal and external validation, respectively. The multi-task-shared encoder was significantly better than the single-task encoder. Furthermore, both fracture and vertebral level detection was good in internal and external validation. Our deep learning model may help radiologists perform real-life medical examinations.http://www.sciencedirect.com/science/article/pii/S2001037023002301Deep LearningComputer-Assisted DiagnosisComputer-Assisted Radiographic Image InterpretationCompression FracturesOsteoporotic Fractures |
| spellingShingle | Seung Min Ryu Soyoung Lee Miso Jang Jung-Min Koh Sung Jin Bae Seong Gyu Jegal Keewon Shin Namkug Kim Diagnosis of osteoporotic vertebral compression fractures and fracture level detection using multitask learning with U-Net in lumbar spine lateral radiographs Computational and Structural Biotechnology Journal Deep Learning Computer-Assisted Diagnosis Computer-Assisted Radiographic Image Interpretation Compression Fractures Osteoporotic Fractures |
| title | Diagnosis of osteoporotic vertebral compression fractures and fracture level detection using multitask learning with U-Net in lumbar spine lateral radiographs |
| title_full | Diagnosis of osteoporotic vertebral compression fractures and fracture level detection using multitask learning with U-Net in lumbar spine lateral radiographs |
| title_fullStr | Diagnosis of osteoporotic vertebral compression fractures and fracture level detection using multitask learning with U-Net in lumbar spine lateral radiographs |
| title_full_unstemmed | Diagnosis of osteoporotic vertebral compression fractures and fracture level detection using multitask learning with U-Net in lumbar spine lateral radiographs |
| title_short | Diagnosis of osteoporotic vertebral compression fractures and fracture level detection using multitask learning with U-Net in lumbar spine lateral radiographs |
| title_sort | diagnosis of osteoporotic vertebral compression fractures and fracture level detection using multitask learning with u net in lumbar spine lateral radiographs |
| topic | Deep Learning Computer-Assisted Diagnosis Computer-Assisted Radiographic Image Interpretation Compression Fractures Osteoporotic Fractures |
| url | http://www.sciencedirect.com/science/article/pii/S2001037023002301 |
| work_keys_str_mv | AT seungminryu diagnosisofosteoporoticvertebralcompressionfracturesandfractureleveldetectionusingmultitasklearningwithunetinlumbarspinelateralradiographs AT soyounglee diagnosisofosteoporoticvertebralcompressionfracturesandfractureleveldetectionusingmultitasklearningwithunetinlumbarspinelateralradiographs AT misojang diagnosisofosteoporoticvertebralcompressionfracturesandfractureleveldetectionusingmultitasklearningwithunetinlumbarspinelateralradiographs AT jungminkoh diagnosisofosteoporoticvertebralcompressionfracturesandfractureleveldetectionusingmultitasklearningwithunetinlumbarspinelateralradiographs AT sungjinbae diagnosisofosteoporoticvertebralcompressionfracturesandfractureleveldetectionusingmultitasklearningwithunetinlumbarspinelateralradiographs AT seonggyujegal diagnosisofosteoporoticvertebralcompressionfracturesandfractureleveldetectionusingmultitasklearningwithunetinlumbarspinelateralradiographs AT keewonshin diagnosisofosteoporoticvertebralcompressionfracturesandfractureleveldetectionusingmultitasklearningwithunetinlumbarspinelateralradiographs AT namkugkim diagnosisofosteoporoticvertebralcompressionfracturesandfractureleveldetectionusingmultitasklearningwithunetinlumbarspinelateralradiographs |