img
img
An Automatic Algorithm For Lung Segmentation and Rib Detection    
Yazarlar (3)
Arş. Gör. Ali ÇINAR Arş. Gör. Ali ÇINAR
Kastamonu Üniversitesi, Türkiye
Bengisu Topuz
Semih Ergin
Eskişehir Osmangazi Üniversitesi, Türkiye
Devamını Göster
Özet
Lung cancer is one of the cancer type which is seen mostly. It appears in spherical form in radiographs. Ribs overlap with the lung tissue so it is important to detect and suppress ribs to see nodules in radiographs more better. The JSRT database is used for this work. It is digital database with and without chest lung nodules. It has 247 images. Firstly, a threshold value is determined by looking all images to eliminate non-lung tissues except ribs outside of the lung. Thus, segmentation of lung is completed. Then, median filter is applied to the images and the difference of median filtered images and segmentated images are obtained to extract rib area. Binarization is applied to the images to see ribs more better. Some lung tissue is connected the ribs so to decrease amount of these tissues, some opening and closing operations are applied to the images. Finally, template matching is applied to the images to get only ribs. Thus, every rib bone’s position are recorded for supressing them later. Totally, two templates are used and these templates are chosen from a random image. One of the rib bones from left side of lung and one of the rib bones from right side of the lung are used as a template. Rib bone template which is chosen from left side of the lung is used to extract rib bones from left side of the lung and rib bone template which is chosen from right side of the lung is used to extract rib bones from right side of the lung. After template matching nonrib tissues are removed and ribs are successfully detected.
Anahtar Kelimeler
Bildiri Türü Tebliğ/Bildiri
Bildiri Alt Türü Tam Metin Olarak Yayınlanan Tebliğ (Uluslararası Kongre/Sempozyum)
Bildiri Niteliği Alanında Hakemli Uluslararası Kongre/Sempozyum
Bildiri Dili İngilizce
Kongre Adı International Conference on Advanced Technology and Sciences (ICAT’xx18)
Kongre Tarihi 28-04-2018 / 01-05-2018
Basıldığı Ülke Türkiye
Basıldığı Şehir Antalya
BM Sürdürülebilir Kalkınma Amaçları
Atıf Sayıları

Paylaş