Due to the characteristics of the optical coherence tomography (OCT) which are non-invasive and harmless for biological tissues, it is possible to observe the inside and outside of blood vessels and the outside of lymphatic vessels. Because the OCT is built based on the interferometer, it uses the reference mirror which causes limitation of imaging speed, so the OCT system can image only narrow region and acquire a few images. To overcome the limitation, the second generation OCT uses wavelength-swept laser and it makes possible to improve imaging speed and image wide region. While the second OCT generates a large number of images, medical doctors or researchers cannot review all blood and lymphatic vessel in images manually. Therefore, many computer-aided studies about providing blood and lymphatic vessel information are published actively, but most of them use two-dimensional structures and simple methods such as filtering. In this thesis, we suggest three methods to detect blood and lymphatic vessel information using three-dimensional (3D) structural information in the second generation blood and lymphatic vessel OCT images. In arterial intravascular OCT images, we find stent region, segment guide-wire, and lumen using 3D structural information, and judge whether stent insertion is right or not. In lymphatic vessel OCT images, we detect lymphatic vessels using 3D information based on 3D modeling and provide quantitative information such as length, surface area, and volume. In retinal blood vessel OCT images, we split the retinal intra-layers using 3D structural information and detect retinal blood vessels using 3D information based on 3D modeling and provide quantitative information.