Lung CT image segmentation algorithm based on region growing and level set method

Abstract

A method is proposed to separate the lung parenchyma from thoracic regions of background noise. Firstly, the traditional region-growing method is applied to locate the contour of the lung. Secondly, the lung boundary noise is removed,the adaptive curvature threshold method is used to repair the lung boundary.Finally, the DRLSE model in the level set method is used to accurately segment the lung regions. Results show that the combining of the two methods for segmenting the lung regions can prevent the edge image missing detection and process different patterns of lung disease images. In a random sample of 150 images, the accuracy of target segmentation can be up to 96.9% and it spents time 0.8 seconds for segmentation an image and the performance shows good robustness. The algorithm can accurately and completely separate the lung region and retain the details of the lung region. It establishes the foundation for the detection and extraction of lung nodules.

Keywords

• region-growing method
• adaptive curvature threshold method
• level set method
• segmenting lung regions