Blurring correction for calculating functional images using an energy resolving photon counting detector

Energy Resolving Photon Counting Detectors (ERPCD) can measure individual X-ray photons with energy discrimination. By analyzing multiple energy images, it can generate functional images that can obtain material information. At the edge of an object image blurring occurs which is caused by the point spread function. Image blurring interferes with the quantification of the object edge in the process of functional image calculation. The purpose of this study is to develop a novel blurring correction method to quantitatively analyze the object edge in functional images. In our correction method, the blurring is determined in a mask area of 5 × 5 pixels. First, in the highenergy image, only the blurred pixels are extracted and corrected by estimating the true value from the mask area. Then, the corresponding pixels in the low-energy image are corrected in the same way. As a result, the blurring is corrected in all pixels of both energy images. To perform theoretical verification of our correction method, a simulation study was conducted. Low- and high-energy images were generated using an aluminum bar phantom having a size of 1 mm × 1 mm and thicknesses of 1 mm to 10 mm taking into consideration detector response, charge sharing, and image blurring. Then, effective atomic number (Z
eff
) images were calculated by analyzing the energy images. When an aluminum bar (Z
eff
= 13) was affected by blurring, proper Z
eff
could not be calculated in the blurred area. However, when applying our method to the blurred image, the blurring was properly corrected at the object edge, which shows the correct value of Z
eff
=13. In addition, we performed experiments using a prototype ERPCD and obtained a similar correction result for a comb-shaped phantom made of aluminum and polymethylmethacrylate. In conclusion, we developed a novel blurring correction method for the quantitative analysis of multiple images.