Artificial images for classifying diffuse lung opacities in thin-section computed tomography images

The classification of diffuse lung opacities in thin-section computed tomography (HRCT) images is very fundamental for developing a computer-aided diagnosis (CAD) system. However, in designing such a CAD system, the number of the available samples is usually small. This leads to the difficulties of designing the CAD system. One way to overcome this problem is to generate artificial images from available real images by image rotation and reversal. In this paper, we discuss the use of artificial images for designing the CAD system.