Cardiovascular disease (CVD) is a fatal disease of the heart or blood vessels. The greatest number of deaths from CVD is coronary heart disease (CHD). It is characterised by thickening of the arterial vessel wall due to atheromatous plaque which may result in narrowing or even occlusion of arterial lumen. Currently, intravascular optical coherence tomography (IVOCT) has been increasingly used in the clinic for the diagnosis of CHD because it permits high-resolution direct tomographic visualisation of cross-sectional images. With IVOCT techniques, stenosis and restenosis caused by plaques and neointima can be detected and analysed. The first main contribution of the thesis is a technique for the automatic segmentation of the lumen border when the guide-wire artifacts are noticeable. The proposed segmentation technique is capable of eliminating guide-wire artifacts and generating accurate lumen borders from IVOCT sequences. Compared to commercially available systems, the proposed method is robust and accurate. The second main contribution of this thesis is an approach for the stent strut detection that can detect stent struts when their intensity responses are weak. This technique is based on stent strut shadow detection. The innovative aspect of our technique is that, for every detected strut shadow, a-priori probability map is applied to estimate the stent strut position. With the detected stent struts, a stent area can be estimated to analyse the neointima hyperplasia (NIH) thickness in IVOCT sequences. The thesis also proposes an approach for the neointima segmentation without any information of the stent but instead with the lumen border. The approach is a combination of a multi-atlas based segmentation approach and a patch-based segmentation approach. With the approach, the neointima label can be obtained by fusing labels from atlases. Compared to other label fusion approaches, a significant increase in segmentation accuracy can be observed.