X-ray spectral computed microtomography (μ-CT) has gained considerable interest in recent years thanks to the development of Photon Counting X-ray Detectors (PCXDs). Through the energy-based discrimination of the photons, obtained by setting up the discriminator threshold at different levels, PCXDs allow measuring photon energies with an accuracy appropriate for several practical applications. The major potentials of this approach are multiple contrast images (sometimes referred to as “color” or “spectral” images), obtained by using a single polychromatic source. In contrast to multiple acquisitions with different tube spectra, photon-counting spectral CT eliminates the risk of misregistration due to motion between consecutive scans and it allows for the elimination of dark noise in the image by rejecting all the events below the signal threshold. Of particular importance is the application of spectral CT in preclinical models where small animals are scanned and an effective discrimination among e.g. soft-tissue, bone and K-edge contrast agents can be in principle performed in a single shot. While spectral CT is considered very promising, its practical application has been hampered by the charge sharing issue that negatively affects the energy resolution of PCXDs. However, latest generations of PCXDs implement solutions to cope with the charge sharing effects, thus allowing sharp color sensitivity. This work presents a comparison of K-edge spectral CT images acquired with the Pixirad-I/Pixie-III detector with and without the application of a charge sharing recovery solution. When a solution for the charge sharing issue is considered both the spatial accuracy and energy resolution are preserved, thus allowing accurate K-edge subtraction imaging.
