To correct a region recognition result in each tomographic image easily. Causing a plurality of tomographic images representing a plurality of regions of a subject or a reconstructed image based on the images and results of recognition processing of the regions of the subject represented by the respective tomographic images, in which the positional relationship of the recognized regions between the tomographic images matches with the anatomical positional relationship of the regions, to be displayed on a screen, accepting input of correction information identifying a correction position which is a boundary of different regions, determining an image whose result of the recognition processing is incorrect and a correct region of the image based on the anatomical positional relationship and/or results of the recognition processing of images adjacent to the correction position, and the correction information, and correcting the result of the recognition processing of the image.