Evaluation of CR Images by Sequential Dependence

Abstract

We described a simple method, i. e, sequential-dependence method, for quantitative evaluation of computed radiographic (CR) images. This method was applied to investigate the effects of image processing which is generally employed in CR systems. Mammograms of breast with microcalcification and chest radiographs processed by gradient enhancement and spatial frequency enhancement techniques, respectively, were used as samples for analysis. Our calculated results indicated that; the higher the gradient used for gradient enhancement the stronger the sequential dependence is, and the lower the spatial frequency used for frequency enhancement the stronger the sequential dependence is. These calculated results are in good agreement with visual results. The preliminary results suggest that the sequential dependence approach is a useful tool for evaluation of overall performance of digital images.