A New Methodology for Blood Vessel Segmentation on Lung CT Images

SAMET R., YILDIRIM Z.

15th Nicograph International Conference (NicoInt), Hangzhou, Çin, 6 - 08 Temmuz 2016, ss.1-7 identifier identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/nicoint.2016.1
  • Basıldığı Şehir: Hangzhou
  • Basıldığı Ülke: Çin
  • Sayfa Sayıları: ss.1-7
  • Anahtar Kelimeler: component, CT, image processing, vessel segmentation, vessel tracking, 3D modeling
  • Ankara Üniversitesi Adresli: Evet

Özet

This paper addresses the problem of blood vessel segmentation of lung on two-dimensional (2D) CT images and three-dimensional (3D) modeling. To this end, a new methodology for blood vessel segmentation is developed. Proposed methodology includes three steps: (1) Lung lobes extraction; (2) Blood vessel segmentation; (3) 3D modeling. Iterative global thresholding, vessel continuity checking and vessel tracking algorithms, developed in the scope of this methodology, are the main contributions of this study. Proposed methodology was implemented on real data and obtained results show that the accuracy of segmented vessels was improved in comparison with existing studies.