| Angiography, the traditional method for evaluating coronary artery disease, is a two-dimensional projection of the complex three-dimensional (3D) coronary artery tree. Only residual lumens can be visualized and no information about the composition or thickness of the vessel wall is available. Intravascular ultrasound (IVUS) provides direct depiction of coronary anatomy and is becoming increasingly accepted as an imaging modality allowing precise tomographic assessment of human coronary anatomy and pathology in vivo. Coronary arteries are interrogated by pullbacks, where the IVUS transducer is slowly withdrawn from the vessel at a constant rate, acquiring tomographic images throughout the length of interest. Quantitative studies on the progression or regression of disease require identification of luminal and medial-adventitial borders in these images. The standard output of IVUS imaging consoles is S-VHS video, which is digitized and then analyzed by manual tracing, a technique that is tedious, time-consuming, and subject to large inter- and intra-observer variability. Automated image processing and analysis techniques can overcome these limitations and make strides toward routine analysis. However, 3D plaque assessment requires techniques that can analyze many images quickly and accurately. In addition, the pulsatile motion of the cardiac cycle introduces cyclic changes in cross-sectional morphometry that make analysis with automated 3D techniques difficult. The goal of this research was to develop and validate a system for the measurement of plaque volume for assessment of lipid-lowering therapy, which required two steps: (1) the development of an electrocardiogram-gated (ECG-gated) acquisition system, which acquires the raw ultrasound backscattered signals at end-diastole, alleviating the cyclic artifacts and bypassing the digitization step, and (2) the development of 3D segmentation and modeling techniques for the identification of luminal and medial-adventitial borders in IVUS image sequences, which allows plaque volume assessment more quickly and reliably than by manual tracing. ECG-gated acquisition and 3D segmentation of IVUS images was tested in a large database of clinical IVUS pullbacks and shown to provide a reliable system for fast assessment of plaque volume in patients undergoing lipid-lowering therapy for coronary atherosclerosis. |
