Introducing a novel image quality measure for digital phase-contrast-image evaluation

Recently, detective quantum efficiency (DQE) arising from the concept of signal-to-noise ratio (SNR) has been used for assessing digital x-ray imaging systems. Using a phase-shift of x-rays that occurs when passing through an object, digital phase contrast imaging (herein referred to as "phase imaging"), which involves magnification, can produce images different from those of standard contact imaging (herein referred to as "regular imaging"). For this reason, assessment of the image quality based on DQE which does not include the object information may not be appropriate to compare image quality between the phase images and the regular images. As an alternative method, we proposed a new image quality assessment method based on radial direction distribution function (RDDF) and signal intensity distribution function (SIDF) in two-dimensional power spectra of images that contain information of an object. To evaluate the usefulness of our method based on RDDF and SIDF, we assessed images of different contrast, noise characteristic and sharpness using simple phantoms. Our results showed that the accurate evaluation of these factors was successfully performed. Comparing the image quality of projected plant seeds by phase imaging and regular imaging, we found the phase imaging method provided higher image quality in terms of edge sharpness than that of the regular imaging.