A New Lung Segmentation Algorithm for Pathological CT Images

This paper presents a new lung segmentation algorithm which is based on anatomical knowledge and Snake model. This algorithm totally overcomes the disadvantage of traditional lung segmentation algorithms, which are mainly based on edge extraction, mathematical morphology, region growing, threshold, etc, and can't get satisfied results when segmenting pathological clinical CT images with traditional algorithms. Experiments showed that no matter whether the CT images are pathological or not, this segmentation algorithm has good results, high speed, and total automation.

[1]  David Zhang,et al.  Extracting the autonomic nerve wreath of iris based on an improved snake approach , 2007, Neurocomputing.

[2]  S. Armato,et al.  Automated lung segmentation for thoracic CT impact on computer-aided diagnosis. , 2004, Academic radiology.

[3]  Daw-Tung Lin,et al.  Lung nodules identification rules extraction with neural fuzzy network , 2002, Proceedings of the 9th International Conference on Neural Information Processing, 2002. ICONIP '02..

[4]  Pheng-Ann Heng,et al.  Research on Boundary Concavities Segmentation via Snake Models , 2005 .

[5]  Demetri Terzopoulos,et al.  Snakes: Active contour models , 2004, International Journal of Computer Vision.

[6]  Rafael Wiemker,et al.  Computer-aided lung nodule detection on high-resolution CT data , 2002, SPIE Medical Imaging.