Objective and Quantitative Evaluation of Image Quality Using Fuzzy Integral: Phantom Study
Abstract
Physical evaluations provide the basis for an objective and quantitative analysis of the image quality. Nonetheless, there are limitations in using physical evaluations to judge the utility of the image quality if the observer's subjectivity plays a key role despite its imprecise and variable nature. This study proposes a new method for objective and quantitative evaluation of image quality to compensate for the demerits of both physical and subjective image quality and combine the merits of them. The images of chest phantom were acquired from four digital radiography systems on clinic sites. The physical image quality was derived from an image analysis algorithm in terms of the contrast-to-noise ratio (CNR) of the low-contrast objects in three regions (lung, heart, and diaphragm) of a digital chest phantom radiograph. For image analysis, various image processing techniques were used such as segmentation, and registration, etc. The subjective image quality was assessed by the ability of the human observer to detect low-contrast objects. Fuzzy integral was used to integrate them. The findings of this study showed that the physical evaluation did not agree with the subjective evaluation. The system with the better performance in physical measurement showed the worse result in subjective evaluation compared to the other system. The proposed protocol is an integral evaluation method of image quality, which includes the properties of both physical and subjective measurement. It may be used as a useful tool in image evaluation of various modalities.
Recommended Citation
S. H. Kim et al., "Objective and Quantitative Evaluation of Image Quality Using Fuzzy Integral: Phantom Study," Korean Journal of Medical Physics, Korean Society of Medical Physics, Jan 2008.
Department(s)
Nuclear Engineering and Radiation Science
Document Type
Article - Journal
Document Version
Citation
File Type
text
Language(s)
English
Rights
© 2008 Korean Society of Medical Physics, All rights reserved.
Publication Date
01 Jan 2008