A Comprehensive Method for Spine Segmentation and Scoliosis Classification Using YOLOv8 and Curvature Analysis

Tinas I., ESMER G. B.

8th IEEE-EMBS Conference on Biomedical Engineering and Sciences, IECBES 2024, Penang, Malezya, 11 - 13 Aralık 2024, ss.58-63, (Tam Metin Bildiri) identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/iecbes61011.2024.10990706
  • Basıldığı Şehir: Penang
  • Basıldığı Ülke: Malezya
  • Sayfa Sayıları: ss.58-63
  • Anahtar Kelimeler: curvature analysis, deep learning, machine learning, medical imaging, Scoliosis, spine segmentation, YOLOv8
  • Yıldız Teknik Üniversitesi Adresli: Evet

Özet

Scoliosis, a medical condition characterized by an abnormal lateral curvature of the spine, presents significant challenges in diagnosis and treatment. This study proposes a comprehensive approach for the segmentation and classification of scoliosis from X-ray images using the YOLOv8 model and curvature analysis. The proposed methodology is based on advanced data preprocessing, spine segmentation using YOLOv8, feature extraction with the Scale-Invariant Feature Transform (SIFT) algorithm, and classification utilizes curvature analysis of the segmented spine. The model achieved high accuracy, outperforming traditional methods and baseline models. The results indicate that our approach provides a robust and efficient solution for scoliosis detection and classification, assisting healthcare professionals in making accurate diagnoses and improving patient outcomes.