New Developments in Intracoronary Ultrasound Processing

New Developments in Intracoronary Ultrasound Processing

Christos V. Bourantas (Michailideion Cardiology Center, Greece & University of Hull, UK), Katerina Naka (Michailideion Cardiology Center, Greece), Dimitrios Fotiadis (Michailideion Cardiology Center, Greece) and Lampros Michalis (Michailideion Cardiology Center, Greece)
DOI: 10.4018/978-1-60566-314-2.ch004
OnDemand PDF Download:
No Current Special Offers


Intracoronary Ultrasound (ICUS) imaging is an intravascular catheter-based technique which provides real-time, high resolution, cross-sectional images of coronary arteries. In these images the lumen, the media-adventitia border, the plaque burden and the composition of the plaque can be identified. Conventionally, ICUS border detection is performed manually. However, this process is laborious and time consuming. To enhance the clinical applicability of ICUS, several automated algorithms have been developed for fast ICUS segmentation and characterisation of the type of the plaque. In this chapter the authors present an overview on the developments in ICUS processing and they describe advanced methodologies which fuse ICUS and X-ray angiographic data in order to overcome indigenous limitations of ICUS imaging and provide complete and geometrically correct coronary reconstruction.
Chapter Preview


Accurate assessment of luminal pathology is useful for the diagnosis and treatment of coronary artery disease. The traditional method used for the depiction of coronary artery morphology is coronary angiography, which provides two–dimensional (2-D) views of the luminal silhouette. Major limitation of this modality is its inability to provide information regarding the plaque burden and the composition of the plaque, data which are useful to guide treatment and estimate prognosis.

To overcome these limitations intracoronary ultrasound (ICUS) has been introduced. ICUS requires the insertion of a catheter, within the coronary artery. At the tip of the catheter there is a transducer which transmits ultrasound signal perpendicular to its axis. There are two types of ICUS systems: the “mechanical” and the “electronic” ICUS systems. In mechanical systems a single rotating transducer, at 1800 rpm (30 revolutions per second), sweeps a high frequency (20 – 40 MHz) ultrasound signal perpendicular to the axis of the catheter, while in electronic systems there is an array of small crystals which have been programmed so that one set of crystals to transmit and the other to receive simultaneously. In both systems cross–sectional images of the coronary artery are produced by detecting the reflections of the ultrasound signal while this is passing through the vessel. As the ICUS catheter is pulled–back (either manually or by a motorized pull–back device) a series of images is generated. In each image, the luminal border, the outer vessel wall border (in the text the term media–adventitia border is used), the stent border, the plaque and the composition of the plaque can be identified and accurate measurements can be obtained (Mintz et al., 2001).

In ICUS images there are often several artefacts which may reduce the ability to identify the regions of interest (Figure 1). These artefacts include: the non–uniform rotational distortion which appears only in the mechanical ICUS systems, the ring down artefact (a bright halo surrounding the transducer) that is due to a high amplitude of the ultrasound signal, the guide wire artefact, the near field artefact, the blood speckles artefact, etc. (Nissen et al., 1993).

Figure 1.

Structures and artefacts observed in an ICUS image


Initially, ICUS border detection was performed manually. However, it became apparent that this process is laborious, time consuming and can be unreliable in the hands of inexperienced operators. Therefore, there was an emerging interest in the development of fast and accurate (semi-) automated segmentation algorithms which would enhance the clinical applicability of ICUS. These algorithms had to face several challenges mostly caused by the high noise content, the low image resolution, the non–uniform luminance and contrast as well as the presence of the above mentioned artefacts.

Another problem that needed to be addressed was the reliable identification of the type of the plaque as well as the integration of the detected ICUS borders into a 3-D object which would represent the coronary vessel. Some of the earlier work on the 3-D reconstruction and visualization of the ICUS sequence assumed that the vessels were straight. However, with this assumption, ICUS could not provide any information on the 3-D arterial geometry or the spatial orientation of the plaque onto the artery. To overcome these limitations fusion of biplane angiographic data and ICUS has been proposed.

In this chapter we attempt to present an overview of the developments in ICUS processing. This review is organised as follows: in the next section we describe the segmentation algorithms which have been introduced for ICUS border detection and plaque characterization. We also present methodologies which have been initially proposed for 3-D coronary reconstruction. In the main part of the chapter we describe novel techniques able to fuse ICUS and angiographic data in order to generate geometrically correct coronary segments. In addition, we present two advanced user–friendly systems, that incorporate these data fusion techniques and allow reliable and comprehensive coronary representation with the general goal to show future trends and interesting potentialities of these systems.

Key Terms in this Chapter

B-splines: Parametrical curves which are described by the following formula:, 4.10where, u is B-spline’s knot vector, Ni,p(u) is B-spline’s basis functions and is the sum of the control points that have been used to approximate spline’s morphology.

Vulnerable Plaque: Atherosclerotic plaque which is prone to thrombosis or has high probability of undergoing rapid progression and causing coronary events.

Neural Network: A modelling technique based on the observed behaviour of biological neurons and used to mimic the performance of a system. It consists of a set of elements (neurons) that start out connected in a random pattern, and, based upon operational feedback, are modelled into the pattern required to generate the optimal results.

Drug Eluting Stent: A metal tube, which is placed into diseased vessels to keep them open and releases a drug that blocks cells’ proliferation

Neointima Coverage: A new or thickened layer of arterial intima formed especially on a prosthesis (e.g. stents) by migration and proliferation of cells from the media.

Optical Coherence Tomography: Intravascular imaging modality which provides cross–sectional images of the vessel. It is similar to ICUS since it requires the insertion of a catheter within the artery, which transmits infrared light. The light is back–reflected as passing through the vessel. These reflections are obtained and analysed to finally generate high resolution cross–sectional images.

Frennet–Serret Formula: A mathematical formula developed by Jean Frédéric Frenet and Joseph Alfred Serret to calculate the curvature and torsion of a 3-D curve.

Complete Chapter List

Search this Book: