Development and clinical evaluation of image quality enhancement techniques for computed radiography systems

In this study, physical and clinical image quality characteristics of low cost and conventional CR systems were evaluated. The image quality control method utilizing a novel phantom was also developed using the contrast-to-noise ratio (CNR) on basis of the observed performances. The baseline image quality of clinical chest computed radiographs was evaluated by for radiologists. Five off-line image post processing techniques were developed using computer MATLAB commands and implemented on radiographs. The results show that the majority of physical and clinical image quality characteristics between low cost and conventional CR systems were comparable. The maximum CNR and contrast detail analysis data on the constructed phantom were obtained at 90 kVp as for other existing phantoms. According to the radiologists, the difference in diagnostic image quality of radiographs of low cost and conventional CR systems were not statistically significant in many cases (0.914 > p > 0.0062) The image quality statistically improved after the implementation of five image post processing techniques, which corresponded to 40% patient dose reduction for one system. The findings from the study have indicated that the physical and clinical performances of low cost CR and conventional CR system are comparable. The concept of CNR is useful in digital radiographs as in mammography when a suitable phantom is used. In addition, suitable off-line image post processing techniques can reduce patient doses while maintaining diagnostic image quality.