Abstract: Objective: To study the application value of radiography, computed tomography (CT) and digital tomosynthesis (DTS) in the screening of novel coronavirus pneumonia (NCP) with chest phantom embedded with ground glass shadow lesion. Methods: A total of 19 Patchy ellipsoid ground glass shadow lesions with image features similar to those of NCP were placed in the lung lobes of the phantom. Low-dose CT, DTS and radiography were respectively used in the inspection of the phantom. After the adjustment of the ground glass shadow lesions' position, the examinations were taken again. The respective detection rate of the three methods, the influence of the location of the chest wall, the location of the lung lobes and size of the ground glass shadow lesions on the detection rate were analyzed. Results: The overall detection rate of CT was 100%, followed by DTS (52.6%) and DR (31.6%), the latter two respectively had statistically significant difference with CT, but there was no statistically significant difference between them. The detection rate of DTS to the ground glass shadow lesions near the lateral chest wall was 100%, but 0 to the ones near the anterior/posterior chest wall. The sensitivity to the size (no less than 2.21cm×1.55cm) was 62.5%, which was not as good as the sensitivity to the anterior-posterior plane position (lateral chest wall) (100%). The minimum size of the ground glass shadow lesions detected by DR in the study was 4.86cm×4.53cm. There was a difference between the size of the ground glass shadow lesions detected and that not detected by DR. The sensitivity of DR to the ground glass shadow lesions size (larger than 4cm×4cm) was 85.7%, which was higher than that (60%) to the anterior-posterior plane position (lateral chest wall). And no significant difference between the detection rates to the ground glass shadow lesions in different lung lobes was found for DTS or DR. The difference of the mean effective dose equivalent values of the three methods was statistically significant. The lowest value was 0.026mSv/time of DR, and the CT was 0.829mSv/time, a little higher than 0.709mSv/time of DTS. Conclusion: Low-dose CT examination is the first choice for NCP patients, except those with severe illness, large inflammatory range and poor mobility, who should be examined by DR technique.