An automatic evaluation system for contrast-detail phantom images in digital radiography

Currently, contrast-detail phantom (CD phantom) images are widely used to assist medical experts for diagnosis, because these images can be used to study imaging quality and assess the radiation dose amount for digital X-ray imaging systems. However, the quality of the CD phantom image varies from person to person, increasing the possibility of false diagnosis. In this article, we propose an automatic evaluation system to measure the quality of a phantom image an alternative to assessment by radiologists. The proposed method is based on bilinear pixel interpolation, geometric transformation, and region shifting to efficiently and objectively evaluate the quality of a phantom image. Moreover, misjudgment modification is also proposed to further improve the recognition accuracy. The experimental results show that the accuracy of the proposed method is very close to that of the radiologists' observations. The results have also proven that the proposed method is a feasible way to evaluate the quality of a phantom image in an objective and automatic manner. (c) 2012 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 22, 214225, 2012 ; Published online in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/ima.22025