[en] Objective: To measure the displacement of solitary pulmonary lesion (SPL) using four-dimensional CT (4DCT), and to compare the planning target volume using 4D maximum intensity projection (MIPMIP) (PTV4D_MIP) with the empirical PTV3D. Methods: Data were acquired from 24 consecutive patients with SPL. For each patient, respiration-synchronized 4DCT images and standard axial CT scans were obtained during free breathing. In lung window setting, the 4D technique was used to measure the displacement of SPL in three dimensions. We compared an PTV created using the MIP (PTV4D_MIP) to the PTV created from the gross tumor volume (GTV) enlarged isotropically for each spatial direction by 1.0 cm and 1. 5 cm in the PTV3D_1.0cm and PTV3D_1.5cm. Results: The SPL located in the lower lobe showed significant difference with the upper and middle lobe in y axis (0. 44 cm, 0.92 cm, t =2.87, P =0.000),but there was no difference in both x and z axis (0.27 cm, 0.39 cm, t =1.44, P =0.116 and 0.29 cm, 0.40 cm, t =1.51, P =0.227). SPL showed significantly greater displacement in y axis than in both x and z axis [0.60 cm and 0.31 cm (t =4.23, P=0.000), 0.60 cm and 0.32 cm (t =4.65, P=0. 000)], but there was no significant difference between x and z axis (0.31 cm, 0.32 cm, t =0.33, P =0.741). There was no statistically difference between the peripheral lung cancer and the pulmonary metastasis tumor in three directions (x axis : 0.37 cm, 0.32 cm, t =0.52, P =0.223; y axis : 0.54 cm, 0.95 cm, t =- 1.38, P =0.061; z axis: 0.42 cm, 0.37 cm, t=0.29, P=0.859). Both PTV3D_1.0cm and PTV3D_1.5cm showed significantly greater volume than PTV4D_MIP (46.73 cm³, 86.52 cm³ and 30.02 cm³, t =-11.35, -12.09, P =0.000, 0.000). Conclusions: The displacement of SPL in y axis is much greater than x and z axis. The empirical PTV3D is much bigger than PTV4D_MIP, which suggests that 4D_MIP provide adequate coverage of the moving target and minimize dose to normal tissues. (authors)