A cascaded dual-pathway residual network for lung nodule segmentation in CT images. (July 2019)

• Record Type: Journal Article
• Authors: Liu, Hong; Cao, Haichao; Song, Enmin; Ma, Guangzhi; Xu, Xiangyang; Jin, Renchao; Jin, Yong; Hung, Chih-Cheng
• Is Part Of: Physica medica. Volume 63(2019)
• Page Start: 112
• Page End: 121
• Publication Date: 2019-07
• Subjects: Lung nodule segmentation -- Residual neural networks -- Cascaded dual-pathway architecture -- Deep learning; Medical physics -- Periodicals; Biophysics -- Periodicals; Biophysics -- Periodicals; Imagerie médicale -- Périodiques; Radiothérapie -- Périodiques; Rayons X -- Sécurité -- Mesures -- Périodiques; Physique -- Périodiques; Médecine -- Périodiques; 610.153
• DOI: 10.1016/j.ejmp.2019.06.003
• Languages: English
• ISSNs: 1120-1797

Abstract

Highlights: For small and juxtapleural nodules, our method can achieve attractive performance. Extract multi-view and multi-scale features of nodules by a cascaded structure. A dual-pathway architecture based on the residual block was proposed. We propose an improved weighted sampling strategy that can adequately sample small nodules. Abstract: It is difficult to obtain an accurate segmentation due to the variety of lung nodules in computed tomography (CT) images. In this study, we propose a data-driven model, called the Cascaded Dual-Pathway Residual Network (CDP-ResNet) to improve the segmentation of lung nodules in the CT images. Our approach incorporates the multi-view and multi-scale features of different nodules from CT images. The proposed residual block based dual-path network extracts local features and rich contextual information of lung nodules. In addition, we designed an improved weighted sampling strategy to select training samples based on the edge. The proposed method was extensively evaluated on an LIDC dataset, which contains 986 nodules. Experimental results show that the CDP-ResNet achieves superior segmentation performance with an average DICE score (standard deviation) of 81.58% (11.05) on the LIDC dataset. Moreover, we compared our results with those of four radiologists on the same dataset. The comparison shows that the CDP-ResNet is slightly better than human experts in terms of segmentation accuracy. Meanwhile, the proposed segmentation method outperforms existing methods.