CLASSIFICATION OF BONE FRACTURES INTO THREE MAIN CLASSES USING CONVOLUTIONAL NEURAL NETWORKS

Creative Commons License

Noğay H. S.

3rd INTERNATIONAL CONGRESS ON ENGINEERING AND SCIENCES, 10 - 11 Mayıs 2024, ss.89-96, (Tam Metin Bildiri)

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Sayfa Sayıları: ss.89-96
  • Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu
  • Bursa Uludağ Üniversitesi Adresli: Evet

Özet

Abstract: Fractures are common injuries with a variety of presentations and clinical outcomes. Accurate classification of bone fractures is crucial to guide treatment decisions and optimize patient outcomes. In this study, we use convolutional neural networks (CNNs) to classify bone fractures into three main classes. The dataset used in this study contains X-ray images of a variety of bone fractures, including avulsion fractures, comminuted fractures, fracture-dislocations, green rod fractures, hairline fractures, impacted fractures, longitudinal fractures, oblique fractures, pathological fractures, and spiral fractures. Each image is labeled according to its corresponding refraction type, yielding ten classes in total. The resulting ten classes are grouped into three basic classes. As a result of the study, an 80% accuracy rate was achieved.