A computer 12 of an image processing apparatus 11 acquires ultrasound B-mode images of consecutive frames to estimate a carotid artery wall at a carotid artery minor axis cross section included in an ultrasound B-mode image of a predetermined frame. The computer 12 also uses, as a template image, the ultrasound B-mode image, which includes the estimated carotid artery wall and surrounding tissues of the carotid artery wall and estimates the size of the diameter of the carotid artery wall so as to minimize a margin of error between a transformed template image, which is created by transforming the template image, and the acquired ultrasound B-mode image of each frame, thereby acquiring the time change of the carotid artery wall.