What is it about?
Computer-aided detection and diagnosis (CAD) of lung-related diseases will be helpful for early detection. Lung parenchyma segmentation is considered as a prerequisite for most of CAD systems. The available traditional methods for lung parenchyma segmentation are not accurate because the nodules that adhere to the lung pleura are recognized as fat. This paper proposes an automated lung parenchyma segmentation for accurate detection of lung nodules, mainly juxtapleural nodules. The proposed method includes the bidirectional chain code to improve the segmentation, and the support vector machine classifier is used to avoid false inclusion of regions. The proposed method is verified on various datasets for robustness of the algorithm. This automated method provides an accuracy of 97% in segmentation compared to ground truth results obtained by experts, which drastically reduces the complexity and intervention of a radiologist.
Featured Image
Why is it important?
Computed tomography (CT) is widely used to diagnose pulmonary parenchyma. CT machines generate more number of slices of volumetric information about nodules with high resolution compared to the traditional chest radiography. This make large data information through single trial and analyse multiple trial data leads to overlook some nodules. In addition, the structural change of nodule detection is more crucial for early detection of cancer. In India, lung cancer is the fourth most frequent cancer, because of its population over one billion. Every year over one million lung cancer patients are registered, in that most of the patients died within the first year. The quality and completeness of the data influence the mortality and incidence rates. In the United States, National Lung Screening Trial (NLST) made the CAD system as a benchmark for lung nodule detection in cancer screening programs, which help 20% reduction in cancermortality due to early diagnosis. These impressive results make country to form Preventive Service Task Force (US-PSTF) to recommended low-dose CT screening annually for adults. This suggestion helps in early detection of cancer, so that many countries recommend the same including India.
Perspectives
The lung parenchyma segmentation is a pre-processing step in computer-aided disease detection and diagnosis. As of now, large numbers of methods are available to segment lung parenchyma. In this paper, the authors’ effort for designing a novel, fully automated, segmentation process using bidirectional chain code and SVM has resulted with good accuracy. Accuracy and robustness of the proposed method are evaluated on large datasets generated by different systems and settings. One hundred and twenty-eight CT studies considered fromLIDC and 52 CT studies considered from RIDER datasets with a total 312 juxtapleural nodules are used in this study. The sample results are verified with ground truth results obtained by two radiologists. The results show that effectiveness of the proposed method is ability to minimize over and under segmentation problems with inclusion of juxtapleural nodules into the lung parenchyma tissue. The proposed method can be used as a pipeline for CAD systems in cancer screening. In real-time screening, some challenges like difficult to identify presence of smaller nodules in low-dose images. This challenge can be comprehensively overcome by using the proposed approach in low-dose studies. In addition, the proposed method is generalized, such that it can be applied to the problems involving convex or concave regions. ACKNOWLEDGMENTS The authors acknowledge the National Cancer Institute and Foundation for the National Institute of Health for critical role in the creation of publicly available LIDC-IDRI and RIDER databases used in this study. Ethical approval: This article does not contain any studies with human participants or animals performed by any of the authors. DISCLOSURE STATEMENT No potential conflict of interest was reported by the authors.
Pramod Kumar
Kalpataru Institute of Technology
Read the Original
This page is a summary of: Lung Parenchyma Segmentation: Fully Automated and Accurate Approach for Thoracic CT Scan Images, IETE Journal of Research, July 2018, Taylor & Francis,
DOI: 10.1080/03772063.2018.1494519.
You can read the full text:
Contributors
The following have contributed to this page
