New Image Processing Models for Opacity Image Analysis in Chest Radiographs

The aim of this research is to develop new image processing models that can be used to characterize single opacities and opacity density within pre-selected regions of interest (ROI). We derive our new image processing models based on the International Labor Organization (ILO) standard for assessing opacity shape and profusion. Pattern spectrum analysis is used for opacity shape and profusion measurements. Opacity enhancement is done using a novel AM-FM model. We use radial pattern spectra to detect the presence of opacities within the ROI, and measure opacity density. The opacity shape and profusion measurement results compare favorably with the ILO standards. The AM-FM model appears to be well-suited for single opacity enhancement.