EUVIMEDEUVIMED entdecken
Open Access

Detection and classification of femoral neck fractures from plain pelvic X-rays using deep learning and machine learning methods

Hüseyin Fatih Sevinç, Kemal Üreten, Talha Karadeniz, Gökhan Koray Gültekin

Turkish Journal of Trauma and Emergency Surgery · 2025-08 · 31 (8) · 783-788

DOI: 10.14744/tjtes.2025.75806 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Turkish Journal of Trauma and Emergency Surgery, 2025-08. Verfasst von Hüseyin Fatih Sevinç und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department Of Orthopedics And Traumatology, Versa Hospital, Kapadokya University, Nevşehir, Turkey. Hauptthemen laut MEDLINE: Femoral Neck Fractures, Deep Learning, Machine Learning, Pelvis. Die Arbeit wurde bisher 4-mal zitiert (Crossref). Der Volltext ist frei in PubMed Central verfügbar.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

BACKGROUND: Femoral neck fractures are a serious health concern, particularly among the elderly. The aim of this study is to diagnose and classify femoral neck fractures from plain pelvic X-rays using deep learning and machine learning algorithms, and to compare the performance of these methods. METHODS: The study was conducted on a total of 598 plain pelvic X-ray images, including 296 patients with femoral neck fractures and 302 individuals without femoral neck fractures. Initially, transfer learning was applied using pre-trained deep learning models: VGG-16, ResNet-50, and MobileNetv2. RESULTS: The pre-trained VGG-16 network demonstrated slightly better performance than ResNet-50 and MobileNetV2 for de-tecting and classifying femoral neck fractures. Using the VGG-16 model, the following results were obtained: 95.6% accuracy, 95.5% sensitivity, 93.3% specificity, 95.7% precision, 95.5% F1 Score, a Cohen's kappa of 0.91, and the Receiver Operating Characteristic (ROC) curve of 0.99. Subsequently, features extracted from the convolution layers of VGG-16 were classified using common machine learning algorithms. Among these, the k-nearest neighbor (k-NN) algorithm outperformed the others and exceeded the accuracy of the VGG-16 model by 1%. CONCLUSION: Successful results were obtained using deep learning and machine learning methods for the detection and clas-sification of femoral neck fractures. The model can be further improved through multi-center studies. The proposed model may be especially useful for physicians working in emergency departments and for those not having sufficient experience in evaluating plain pelvic radiographs.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Femoral Neck FracturesDeep LearningMachine LearningPelvisHumansRadiographyMaleAgedFemaleMiddle AgedAlgorithmsSensitivity and SpecificityROC Curve

Autor:innen und Institutionen

  • Department Of Orthopedics And Traumatology, Versa Hospital, Kapadokya University, Nevşehir, Turkey.
  • Department Of Computer Engineering, Çankaya University, Ankara, Turkey.
  • Department Of Electrical And Electronics Engineering, Yıldırım Beyazıt University, Ankara, Turkey.

Zitieren

APA

Sevinç, H. F., Üreten, K., Karadeniz, T., & Gültekin, G. K. (2025). Detection and classification of femoral neck fractures from plain pelvic X-rays using deep learning and machine learning methods. Turkish Journal of Trauma and Emergency Surgery, 31(8), 783-788. https://doi.org/10.14744/tjtes.2025.75806

Vancouver

Sevinç HF, Üreten K, Karadeniz T, Gültekin GK. Detection and classification of femoral neck fractures from plain pelvic X-rays using deep learning and machine learning methods. Turkish Journal of Trauma and Emergency Surgery. 2025;31(8):783-788. doi:10.14744/tjtes.2025.75806

BibTeX
@article{Sevin2025,
  author = {Sevinç, Hüseyin Fatih and Üreten, Kemal and Karadeniz, Talha and Gültekin, Gökhan Koray},
  title = {Detection and classification of femoral neck fractures from plain pelvic X-rays using deep learning and machine learning methods},
  journal = {Turkish Journal of Trauma and Emergency Surgery},
  year = {2025},
  volume = {31},
  number = {8},
  pages = {783--788},
  doi = {10.14744/tjtes.2025.75806}
}
Kostenlos starten →