Integrated approach for classification and segmentation of lung anatomical structures from chest ct images
Loading...
Date
item.page.authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Computed Tomography (CT) is the best imaging technique used for the detection and diagnosis of the cancerous nodules, due to its high spatial resolution, fast imaging speed and wide availability. Nodules of the lungs and pathological residues with varied diameter can be comfortably viewed by computed tomography and can be categorized as benign or malignant. The key intention of this segmentation and smoothing is to develop an efficient methodology for an automated lung tumour diagnosis. Segmentation is the quantitative pre- processing step used in the chest CT analysis. The iterative weighted averaging technique is used in addressing the issues related to the segmentation and smoothing method. The methodology incorporates the accurate Lung Segmentation, enclosure of Juxtapleural nodules and proper insertion of the bronchial vessels for enhancing the smoothness of the segmented Lung area. The steps involved in this research include Image preprocessing by nonlinear anisotropic diffusion filtering, Thorax Extraction, Lung segmentation and iterative weighted averaging technique for smoothing the contours. The main feature in choosing the nonlinear anisotropic diffusion filtering is for properly preserving the regions unaffected with cancer and also to differentiate the artefacts involved due to the subject movements. In the fuzzy c- means clustering algorithm, the lung parenchyma is identified from the thorax region based on the fuzzy membership value and the affected regions are attached. The main advantage of fuzzy c- means clustering is that it allows continuing membership of data points to clusters measured as degrees in [0,1]. This gives the flexibility to express that data points can belong to more than one cluster.
newline