An imaging apparatus is disclosed for diagnosis which includes first generating means for generating a closed curve indicating a blood vessel wall by extracting a position of the blood vessel wall using an ultrasonic cross-sectional image at a predetermined position in an axial direction, second generating means for generating a closed curve indicating a boundary of a blood flow region by extracting a boundary position of the blood flow region using an optical cross-sectional image at the predetermined position, and setting means for setting a region between the closed curve generated by the first generating means and the closed curve generated by the second generating means as the region of interest in the ultrasonic cross-sectional image or the optical cross-sectional image, in a case where the ultrasonic cross-sectional image and the optical cross-sectional image are aligned and overlapped with each other at the predetermined position.