New Method Improves Image Quality in Spectral Photon Counting CT Imaging
Former OMA Fellow Navrit Bal and colleagues from Aarhus University have introduced a novel post processing method to enhance the quality of spectral photon counting cone beam computed tomography (CT) imaging, which is often challenged by pixel response inconsistencies and resulting image artefacts, such as rings. The new approach, called digital-to-analogue converter (DAC-)shifting, explicitly addresses these pixel response issues, leading to significant improvements in image clarity and accuracy. Spectral photon counting CT imaging, which captures detailed images by detecting X ray photons at different energy levels, can suffer from noise and artefacts due to variations in individual pixel responses. Existing methods to correct for these issues typically rely on calibration techniques, such as signal to thickness calibration or post processing corrections that do not directly account for pixel response.
The DAC shifting method, as presented in a recent study, measures pixel responses using flat field images and then adjusts the projection data to correct these inconsistencies. The technique was evaluated using a series of spectral photon counting CT scans of a phantom with varying density inserts and iodine K edge imaging. Results showed that DAC shifting was highly effective in reducing CT number variation in homogeneous materials, achieving an average reduction of 47.4% compared to the more limited 13.7% reduction seen with STC.
In addition to improving image uniformity, DAC shifting enhanced iodine K edge imaging by producing sharper attenuation peaks and more consistent CT values. This could have important implications for more accurate iodine concentration quantification in clinical applications, such as cancer imaging.
The findings suggest that DAC shifting is a robust and effective method for improving spectral photon counting CT imaging, with potential for widespread use in both research and clinical settings.
Full article:
Bal, Navrit Johan Singh; Ragupathy , Imaiyan Chitra ; Tramm , Trine; Nijkamp, Jasper , “A Novel and Reliable Pixel Response Correction Method (DAC Shifting) for Spectral Photon Counting CT Imaging”, TOMOGRAPHY 10(7), 1168 1191 (JUL 2024)