It has been difficult to display medical three-dimensional images taken by different imaging devices (imaging methods or modalities) individually in a common display direction. In the invention, a coordinate system 1 on the basis of landmarks on the bones and a coordinate system 2 on the basis of landmarks on the skins are set in a CT image, and a transformation coefficient (correction amount) Tc between the coordinate system 1 and the coordinate system 2 is calculated and stored. In a case where image data acquired by another imaging method is displayed, a coordinate system same as the coordinate system 1 in the CT image can be set in a case where the image data includes a bone display layer, and a coordinate system same as the coordinate system 2 in the CT image can be set in a case where the image displays a skin display layer. By correcting the coordinate system thus set using the transformation coefficient Tc, the coordinate system 1 can be corrected to the coordinate system 2 or the coordinate system 2 to the coordinate system 1. It is thus possible to display images according to the common coordinate system between different sets of image data.