Deep-learning-based pelvic automatic segmentation in pelvic fractures.
Deep learning
Pelvic fracture
Segmentation
Journal
Scientific reports
ISSN: 2045-2322
Titre abrégé: Sci Rep
Pays: England
ID NLM: 101563288
Informations de publication
Date de publication:
28 May 2024
28 May 2024
Historique:
received:
12
12
2023
accepted:
24
05
2024
medline:
29
5
2024
pubmed:
29
5
2024
entrez:
28
5
2024
Statut:
epublish
Résumé
With the recent increase in traffic accidents, pelvic fractures are increasing, second only to skull fractures, in terms of mortality and risk of complications. Research is actively being conducted on the treatment of intra-abdominal bleeding, the primary cause of death related to pelvic fractures. Considerable preliminary research has also been performed on segmenting tumors and organs. However, studies on clinically useful algorithms for bone and pelvic segmentation, based on developed models, are limited. In this study, we explored the potential of deep-learning models presented in previous studies to accurately segment pelvic regions in X-ray images. Data were collected from X-ray images of 940 patients aged 18 or older at Gachon University Gil Hospital from January 2015 to December 2022. To segment the pelvis, Attention U-Net, Swin U-Net, and U-Net were trained, thereby comparing and analyzing the results using five-fold cross-validation. The Swin U-Net model displayed relatively high performance compared to Attention U-Net and U-Net models, achieving an average sensitivity, specificity, accuracy, and dice similarity coefficient of 96.77%, of 98.50%, 98.03%, and 96.32%, respectively.
Identifiants
pubmed: 38806582
doi: 10.1038/s41598-024-63093-w
pii: 10.1038/s41598-024-63093-w
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
12258Informations de copyright
© 2024. The Author(s).
Références
Kim, K. H., Kim, S. J. & Sung, C. S. A clinical study on fractures of the pelvic bone. J. Korean Orthop. Assoc. 17, 485–491 (1982).
doi: 10.4055/jkoa.1982.17.3.485
Park, K.-C. Pelvic fracture issues. J. Korean Fract. Soc. 23, 341–345 (2010).
doi: 10.12671/jkfs.2010.23.3.341
Kim, B. H., Im, J. I., Kim, D. J. & Park, J. Y. Diagnostic values of bone scan followed by CT scan in undetected pelvic bone fracture. J. Korean Orthop. Assoc. 32, 523–529 (1997).
doi: 10.4055/jkoa.1997.32.3.523
Kim, S. Y., Park, B.-H.P. & Ihn, J.-C. Diagnostic value of computed tomography in acetabular fracture. J. Yeungnam Med. Sci. 5, 43–48 (1988).
doi: 10.12701/yujm.1988.5.1.43
Han, H.-J. Diagnosis of pelvic bone fractures. J. Korean Fract Soc. 7, 5–11 (1994).
doi: 10.12671/jksf.1994.7.1.5
Ho, L., Dong-seong, K. & Heung-sik, K. Bone segmentation method of visible human using multimodal registration. J. KIISE Softw. Appl. 30, 719–726 (2003).
Min-Jun, S. & Do-Yeon, K. Pulmonary vascular segmentation and refinement on the CT scans. J. Korea Inst. Inf. Commun. Eng. 16, 591–597 (2012).
Kim, Y. J., Park, S. J., Kim, K. R. & Kim, K. G. Automated ulna and radius segmentation model based on deep learning on DEXA. J. Korea Multimed. Soc. 21, 1407–1416 (2018).
Wang, Y. et al. Weakly supervised universal fracture detection in pelvic x-rays in Med. In Image Comput. Comput. Assist. Interv. MICCAI. Proceedings of the Part VI: 22nd International Conference, Shenzhen, China, October 13–17, vol. 22 459–467 (Springer, 2019).
Liu, X. et al. Fully automated pelvic bone segmentation in multiparameteric MRI using a 3D convolutional neural network. Insights Imaging 12, 93 (2021).
doi: 10.1186/s13244-021-01044-z
pubmed: 34232404
pmcid: 8263843
Yuan, S. et al. Rethinking computer-aided pelvis segmentation. In ICASSP IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 451–1455 (IEEE, 2022).
Wu, J. et al. U-Net combined with multi-scale attention mechanism for liver segmentation in CT images. BMC Med. Inform. Decis. Mak. 21, 283 (2021).
doi: 10.1186/s12911-021-01649-w
pubmed: 34654419
pmcid: 8520298
Noguchi, S., Nishio, M., Yakami, M., Nakagomi, K. & Togashi, K. Bone segmentation on whole-body CT using convolutional neural network with novel data augmentation techniques. Comput. Biol. Med. 121, 103767 (2020).
doi: 10.1016/j.compbiomed.2020.103767
pubmed: 32339097
Wasserthal, J. et al. Totalsegmentator: Robust segmentation of 104 anatomic structures in ct images. Rad. Artif. Intell. 5, 5 (2023).
Kai, H., Feng, Z. Y., Meng, H., Baoping, F. Y. & Han, Y. R. Ultrasound image segmentation of breast tumors based on Swin-transformerv2. In Proceedings of the 2022 10th International Conference on Information Technology: IoT and SMART City 106–111 (2022).
Gao, Z.-J., He, Y. & Li, Y. A novel lightweight Swin-unet network for semantic segmentation of COVID-19 lesion in CT images. IEEE Access 11, 950–962 (2022).
doi: 10.1109/ACCESS.2022.3232721
Urakawa, T. et al. Detecting intertrochanteric hip fractures with orthopedist-level accuracy using a deep convolutional neural network. Skelet. Rad. 48, 239–244 (2019).
doi: 10.1007/s00256-018-3016-3
Lee, C. G., Kwon, J. W., Sohn, K. T., Shin, S. H. & Park, J. W. Combined injuries of pubic ramus fracture: The role of computerized tomography. J. Korean Soc. Fract. 12, 40–46 (1999).
doi: 10.12671/jksf.1999.12.1.40
Popovic, A., de la Fuente, M., Engelhardt, M. & Radermacher, K. Statistical validation metric for accuracy assessment in medical image segmentation. Int. J. Comput. Assist. Rad. Surg. 2, 169–181 (2007).
doi: 10.1007/s11548-007-0125-1
Jois, P. S., Manjunath, A. & Fevens, T. Boosting segmentation performance across datasets using histogram specification with application to pelvic bone segmentation. In IEEE International Conference on Bioinformatics and Biomedicine (BIBM) 1364–1369 (IEEE, 2021).
Jodeiri, A. et al. Region-based convolution neural network approach for accurate segmentation of pelvic radiograph. In 26th National and 4th International Iranian Conference on Biomedical Engineering (ICBME) 152–157 (IEEE, 2019).