KAFATASI KIRIKLARININ TEŞHİSİNDE YAPAY ZEKÂ: GÜNCEL DERİN ÖĞRENME MİMARİ PERFORMANSLARI

Fidan Kaya Gülağız; Tuğrul Hakan Gençtürk; İsmail Kaya

doi:10.17780/ksujes.1515526

TR EN

KAFATASI KIRIKLARININ TEŞHİSİNDE YAPAY ZEKÂ: GÜNCEL DERİN ÖĞRENME MİMARİ PERFORMANSLARI

Öz

Kafa travmaları, ciddi sonuçlara yol açabilen ve etkileri uzun yıllar sürebilecek sağlık sorunlarından biridir. Teşhis, ilk aşamada nörolojik muayene ile başlar ve gerektiğinde bilgisayarlı tomografi (BT) kullanılır. Kafatası kırıkları, diğer kafa travmalarına göre daha ciddi hasarlara eşlik ederler ve sıkça görülür. Özellikle ilk müdahalenin pratisyen hekimler ve acil uzmanları tarafından yapılması, BT görüntülerinin yorumlanmasında uzmanlık ve destek gerektirir. Bu noktada, özellikle ilk teşhis ve tanı aşamasında hekimlere destek olacak yapay zeka tekniklerinin varlığı büyük bir önem taşımaktadır. Bu çalışmada kafatası kırığının tespiti için kullanılabilecek dört farklı mimarinin alt modelleriyle birlikte kapsamlı bir karşılaştırması yapılmıştır. Bu amaçla Verimli Sinir Ağı (EfficientNet), Artık Ağlar (ResNet), Residual Networks with Aggregated Residual Transformations (ResNeXt) ve Maximum Vision Transformer (MaxVit) mimarileri çalışmaya dahil edilmiştir. Modellerin kafatası kırığını sınıflandırma açısından başarısı çalışmaya özgü olarak toplanan kapsamlı ve güncel bir veri kümesi üzerinden gösterilmiştir. Deneysel sonuçlar ile hem hangi yöntemin kafatası kırığı açısından daha uygun ve doğru sonuçlar verdiği ortaya konulmuş hem de güncel derin öğrenme mimarilerinin bu alandaki durumu özetlenmiştir.

Anahtar Kelimeler

ARTIFICIAL INTELLIGENCE IN SKULL FRACTURE DIAGNOSIS: CURRENT DEEP LEARNING ARCHITECTURE PERFORMANCES

Abstract

Cranial injuries are one of the health issues that can lead to serious consequences and have effects that may last for many years. Diagnosis begins with a neurological examination in the initial stage and employs computed tomography (CT) when necessary. Skull fractures are more frequently accompanied by severe injuries compared to other types of head trauma and are commonly observed. The initial intervention is often performed by general practitioners and emergency specialists, necessitating expertise and support in interpreting CT images. At this point, the presence of artificial intelligence techniques to assist physicians, particularly during the initial diagnosis and evaluation stages, is of significant importance. In this study, a comprehensive comparison of four different architectures that can be used for skull fracture detection has been made together with their sub-models. For this purpose, Efficient Neural Network (EfficientNet), Residual Network (ResNet), Residual Networks with Aggregated Residual Transformations (ResNeXt) and Maximum Vision Transformer (MaxVit) architectures are included in the study. The success of the models in terms of skull fracture classification is demonstrated on a comprehensive and up-to-date dataset collected specifically for the study. With the experimental results, both which method is more appropriate and accurate for skull fracture classification and the state of the art of current deep learning architectures in this field are summarized

Keywords

Kaynakça

Abubacker, N. F., Azman, A., Azrifah, M., & Doraisamy, S. (2013, December). An approach for an automatic fracture detection of skull dicom images based on neighboring pixels. In 2013 13th International Conference on Intellient Systems Design and Applications (pp. 177-181). IEEE. https://doi.org/10.1109/ISDA.2013.6920731
Bakchy, S. C., Peyal, H. I., Islam, M. I., Yeamin, G. K., Miraz, S., & Abdal, M. N. (2023, September). A lightweight-CNN model for efficient lung cancer detection and Grad-CAM visualization. In 2023 International Conference on Information and Communication Technology for Sustainable Development (ICICT4SD) (pp. 254-258). IEEE. https://doi.org/10.1109/ICICT4SD59951.2023.10303569 Brain Trauma Foundation. Traumatic brain injury statistics. (2024). https://www.braintrauma.org/ Son Erişim: 01.07.2024
Chilamkurthy, S., Ghosh, R., Tanamala, S., Biviji, M., Campeau, N. G., Venugopal, V. K., ... & Warier, P. (2018). Deep learning algorithms for detection of critical findings in head CT scans: a retrospective study. The Lancet, 392(10162), 2388-2396. https://doi.org/10.1016/S0140-6736(18)31645-3
Choi, J. W., Cho, Y. J. , Ha, J. Y., Lee, Y. Y. , Koh, S. Y. , Seo, J. Y. , Choi, Y. H. , Cheon, J. , Phi, J. H. , Kim, I. , Yang, J. & Kim, W. S. (2022). Deep learning-assisted diagnosis of pediatric skull fractures on plain radiographs. Korean Journal of Radiology, 23(3), 343-354. https://doi.org/10.3348/kjr.2021.0449.
Gençtürk, T. H., Gülağız, F. K., & Kaya, İ. (2023). Derin öğrenme yöntemleri kullanılarak BT taramalarında beyin kanaması teşhisinin karşılaştırmalı bir analizi. Journal of Intelligent Systems: Theory and Applications, 6(1), 75-84. https://doi.org/10.38016/jista.1215025
Gençtürk, T. H., GülağIz, F. K., & Kaya, İ. (2024). Detection and segmentation of subdural hemorrhage on head CT images. IEEE Access, 12, 82235-82246. https://doi.org/10.1109/ACCESS.2024.3411932 Greenberg, M. S. (2010). Handbook of neurosurgery. (7th ed.) Thieme. ISBN: 978-1-60406-326-4.
Guo, Y., He, Y., Lyu, J., Zhou, Z., Yang, D., Ma, L., ... & Dai, Q. (2022). Deep learning with weak annotation from diagnosis reports for detection of multiple head disorders: a prospective, multicentre study. The Lancet Digital Health, 4(8), e584-e593. https://doi.org/10.1016/S2589-7500(22)00090-5
Han, J., Kamber, M., & Pei, J. (2012). Data mining concepts and techniques. (3rd ed.). Waltham: Morgan Kaufmann Publishers.

He, K., Zhang, X., Ren, S., & Sun, J. (2016, June). Deep residual learning for image recognition. In Proceedings of the IEEE Conference On Computer Vision And Pattern Recognition (pp. 770-778). IEEE. https://doi.org/10.1109/CVPR.2016.90
Kaya, İ., Gençtürk, T. H. & Kaya Gülağız, F. (2023). A revolutionary acute subdural hematoma detection based on two-tiered artificial intelligence model. Ulusal Travma Acil Cerrahi Dergisi, 29(8), 858–871. PMID: 37563894; PMCID: PMC10560802. https://doi.org/10.14744/tjtes.2023.76756
Lin, X. , Yan, Z. , Kuang, Z., Zhang, H., Deng, X. & Yu L. (2022). Fracture R‐CNN: An anchor‐efficient anti‐interference framework for skull fracture detection in CT images. Medical Physics, 49(11), 7179-7192. https://doi.org/10.1002/mp.15809
Lin, E., & Yuh, E. L. (2022). Computational approaches for acute traumatic brain injury image recognition. Frontiers in Neurology, 13, 791816, 1-25. https://doi.org/10.3389/fneur.2022.791816
Maconochie, I. & Ross, M. (2007). Head injury (moderate to severe). BMJ Clin Evid, 2007(1210), 1-13, 2007. PMID: 19450357; PMCID: PMC2943769.
Mangrulkar, A., Rane, S. B., & Sunnapwar, V. (2021). Automated skull damage detection from assembled skull model using computer vision and machine learning. International Journal of Information Technology, 13, 1785-1790. https://doi.org/10.1007/s41870-021-00752-5
Muschelli III, J. (2020). ROC and AUC with a binary predictor: a potentially misleading metric. Journal of Classification, 37(3), 696-708. https://doi.org/10.1007/s00357-019-09345-1
Sehlikoğlu, K., Türkoğlu, A., Bork, T., & Batbaş, M. (2024). Investigation of fatal traumatic head injuries. Ulus Travma Acil Cerrahi Derg, 30(3), 160-166. PMID: 38506383; PMCID: PMC10977496. https://doi.org/10.14744/tjtes.2024.32463
Selvaraju, R. R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., & Batra, D. (2017). Grad-cam: Visual explanations from deep networks via gradient-based localization. In Proceedings Of The IEEE International Conference On Computer Vision (pp. 618-626). IEEE. https://doi.org/10.1109/ICCV.2017.74
Shan, W., Guo, J., Mao, X., Zhang, Y., Huang, Y., Wang, S., ... & Wang, Y. (2021). Automated identification of skull fractures with deep learning: a comparison between object detection and segmentation approach. Frontiers in Neurology, 12, 687931, 1-10. https://doi.org/10.3389/fneur.2021.687931
Shao, H., & Zhao, H. (2003, September). Automatic analysis of a skull fracture based on image content. In Third International Symposium on Multispectral Image Processing and Pattern Recognition (pp. 741-746). SPIE. https://doi.org/10.1117/12.538780
Tan, M., & Le, Q. (2019, May). Efficientnet: Rethinking model scaling for convolutional neural networks. In International Conference On Machine Learning (pp. 6105-6114). PMLR. https://doi.org/10.48550/arXiv.1905.11946
Tu, Z., Talebi, H., Zhang, H., Yang, F., Milanfar, P., Bovik, A., & Li, Y. (2022, October). MaxVit: Multi-axis vision transformer. In European Conference on Computer Vision (pp. 459-479). Cham: Springer Nature Switzerland. https://doi.org/10.48550/arXiv.2204.01697
Türkiye Cumhuriyeti Sağlık Bakanlığı. Health statistics Türkiye, Health statistics yearbook for Türkiye . (2019). https://www.saglik.gov.tr/TR,84966/saglik-istatistikleri-yilligi-2019-yayinlanmistir.html Son Erişim: 23.06.2024.
Türkiye Ministry of Health Expert Board in Medicine. Türkiye emergency medicine specialty training curriculum. (2022). https://tuk.saglik.gov.tr›aciltipmufredatv24doc. Son Erişim: 23.06.2024.
Wallis, A. & McCoubrie, P. (2011). The radiology report--are we getting the message across?. Clin Radiol, 66(11), 1015-1022. https://doi.org/10.1016/j.crad.2011.05.013
Wan Zaki, W. M. D., Ahmad Fauzi, M. F., & Besar, R. (2009, November). A new approach of skull fracture detection in CT brain images. In Visual Informatics: Bridging Research and Practice: First International Visual Informatics Conference (pp. 156-167). Springer Berlin Heidelberg.
Wang, H. C., Wang, S. C., Yan, J. L., & Ko, L. W. (2023). Artificial Intelligence model trained with sparse data to detect facial and cranial bone fractures from head CT. Journal of Digital Imaging, 36(4), 1408-1418. https://doi.org/10.1007/s10278-023-00829-6
Xie, S., Girshick, R., Dollár, P., Tu, Z., & He, K. (2017, July). Aggregated residual transformations for deep neural networks. In Proceedings of the IEEE Conference On Computer Vision And Pattern Recognition (pp. 1492-1500). IEEE. https://doi.org/10.1109/CVPR.2017.634
Yamada, A., Teramoto, A., Otsuka, T., Kudo, K., Anno, H., & Fujita, H. (2016, August). Preliminary study on the automated skull fracture detection in CT images using black-hat transform. In 2016 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) (pp. 6437-6440). IEEE. https://doi.org/10.1109/EMBC.2016.7592202
Yamada, A., Teramoto, A., Kudo, K., Otsuka, T., Anno, H., & Fujita, H. (2018). Basic study on the automated detection method of skull fracture in head ct images using surface selective black-hat transform. Journal of Medical Imaging and Health Informatics, 8(5), 1069-1076. https://doi.org/10.1166/jmihi.2018.2410
Zaki, W. M. D. W. , Fauzi, M. F. A. & R. Besar. (2008, November) .Automated method of fracture detection in CT brain images. In 2008 3rd International Conference on Intelligent System and Knowledge Engineering (pp. 1156-1160). IEEE. https://doi.org/10.1109/ISKE.2008.4731105

Ayrıntılar

Birincil Dil

Türkçe

Konular

Derin Öğrenme

Bölüm

Araştırma Makalesi

Yazarlar

Fidan Kaya Gülağız ^*
0000-0003-3519-9278
Türkiye

Tuğrul Hakan Gençtürk
0000-0002-2736-271X
Türkiye

İsmail Kaya
0000-0002-4128-5845
Türkiye

Yayımlanma Tarihi

3 Mart 2025

Gönderilme Tarihi

12 Temmuz 2024

Kabul Tarihi

4 Ekim 2024

Yayımlandığı Sayı

Yıl 2025 Cilt: 28 Sayı: 1

DOI

https://doi.org/10.17780/ksujes.1515526

IZ

https://izlik.org/JA46BE85GY

Kaynak Göster

RIS / Bibtex

APA

Kaya Gülağız, F., Gençtürk, T. H., & Kaya, İ. (2025). KAFATASI KIRIKLARININ TEŞHİSİNDE YAPAY ZEKÂ: GÜNCEL DERİN ÖĞRENME MİMARİ PERFORMANSLARI. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi, 28(1), 51-64. https://doi.org/10.17780/ksujes.1515526