ObjectiveTo explore the impact of different monochromatic reconstruction on image quality of early lesions of coronavirus disease 2019 (COVID-19).MethodsThe chest spectral CT images of 11 patients confirmed as COVID-19 in West China Hospital of Sichuan University between January and February 2020 were retrospectively analyzed. A total of 34 inflammatory lesions were found in the 11 cases. Seven groups of images were reconstructed from the raw data for each patient, including the conventional polychromatic image and different monochromatic images of 40-140 keV (with intervals of 20 keV). CT and standard deviation (SD) values of all lesions were measured to calculate the signal-noise ratio (SNR) and contrast-noise ratio (CNR). The image quality was subjectively scored by two radiologists, and the differences in image quality among different monochromatic groups and the polychromatic group were compared.ResultsWith the increase of X-ray energy, the CT values and SD values of reconstructed images in monochromatic groups gradually decreased, and the SNRs and CNRs gradually increased, and the differences between adjacent two groups were all statistically significant (P<0.001). In the range of 80-140 keV, the SD values of different monochromatic groups were lower than that of the polychromatic group, and the SNRs and CNRs were higher than those of the polychromatic group, and the pairwise comparison results showed statistically significant differences (P<0.001). The 120 keV-reconstructed image had the highest subjective score, and the difference from that of the polychromatic image was statistically significant (P<0.05).ConclusionsDifferent monochromatic reconstruction of spectral CT can significantly reduce the image noise in early COVID-19 lesions, and improve the image quality. Combining subjective and objective evaluation of images, the 120 keV-reconstructed monochromatic image shows the best early lesions of COVID-19 and is of great significance for early clinical screening.
