Performance and Analysis of Automatic Detection Of Ground-Glass Pattern in Lung Disease using High-Resolution Computed Tomography

Abstract

This study proposes an approach for automatic detection of Ground glass pattern, a lung disease, from Computed Tomography (CT) and High Resolution Computed Tomography (HRCT) scans of the lung. The algorithm is based on frequency spectrum analysis of image using Gabor filter bank. Gabor filter banks are used to support the frequency extraction process. These algorithms when applied to HRCT images will assist doctors to gain more information than from the CT images. The tasks are completed in three steps: Preliminary mask formation, Peripheral mask formation and finally post processing. By these, higher sensitivity and selectivity may be achieved with fast processing time. In the post processing, binary noise removal technique is used to remove noise from the detection mask.