Abstract—The classification of diffuse lung opacities in thin-section computed tomography (HRCT) images is an important step for developing a computer-aided diagnosis (CAD) system. In designing the CAD system for classifying diffuse lung opacities in HRCT images, a local histogram feature vector approach has been proposed and shown to be effective. Furthermore, a local slice feature vector approach unlike a local histogram feature approach has been proposed and shown to be effective. However, the effectiveness of a combined feature vector of local histogram and local slice feature vectors is not clear. In this paper, more effectively to classify diffuse lung opacities in HRCT images, we explore the combined feature vector of local histogram and local slice feature vectors. Furthermore, we examine the effects of normalization, rescaling and standardization, for the combined feature vector.
Index Terms—Diffuse lung opacities, combined feature vector, local histogram feature, local slice feature, HRCT images, CAD system.
Yoshihiro Mitani is with the National Institute of Technology, Ube College, Ube, Japan (e-mail: mitani@ ube-k.ac.jp).
Yusuke Fujita, Naofumi Matsunaga, and Yoshihiko Hamamoto are with Yamaguchi University, Ube, Japan.
[PDF]
Cite:Yoshihiro Mitani, Yusuke Fujita, Naofumi Matsunaga, and Yoshihiko Hamamoto, "Combining of Local Histogram and Local Slice Feature Vectors for Classifying Diffuse Lung Opacities in Thin-Section Computed Tomography Images," International Journal of Information and Electronics Engineering vol. 6, no. 2, pp. 135-138, 2016.
