Automatic segmentation for pulmonary nodules in CT images based on multifractal analysis

Abstract

To characterise pulmonary nodules, the volume analysis of solitary pulmonary nodules (SPNs) is important for diagnosis. To accurately estimate the volume of pulmonary nodules and reduce physician workloads, this study presents an intelligent method to automatically detect and segment pulmonary nodules from computed tomography (CT) scans. To detect nodule candidates, a modified ‘Iso’ capacity measure based on multifractals analysis is proposed to reflect the local singularity properties and capture the textures of pulmonary nodules. Subsequently, the proposed two-stage false nodules pruning procedure is applied to categorise a true SPN from nodule candidates. Preliminary results demonstrate that the proposed method can automatically and successfully detect and segment 118 SPNs in 118 CT images from a public lung database. The average and standard deviation of segmentation overlap measures are 0.71 and 0.08, respectively. The authors’ method is competitive compared to the methods reported in the literature. Besides, the proposed method can avoid the problem of selecting seeds when applying region-growing techniques to segment pulmonary nodules.