Toward Novel Noninvasive and Low-Cost Markers for Predicting Strokes in Asymptomatic Carotid Atherosclerosis: The Role of Ultrasound Image Analysis

Abstract

Stroke is a serious and frequent cerebrovascular disease with an enormous socioeconomic burden worldwide. Stroke prevention includes treatment of carotid atherosclerosis, the most common underlying cause of stroke, according to a specific diagnostic algorithm. However, this diagnostic algorithm has proved insufficient for a large number of mostly asymptomatic subjects, which poses a significant research challenge of identifying novel personalized risk markers for the disease. This paper illustrates the potential of carotid ultrasound image analysis toward this direction, with ultrasound imaging being a low-cost and noninvasive imaging modality and ultrasound-image-based features revealing valuable information on plaque composition and stability. A concise report of state-of-the-art studies in the field is provided and a perspective for clinical scenario for optimal management of atherosclerotic patients is described. Challenges and necessary future steps toward the realization of this scenario are discussed in an attempt to urge and orient future research, and mainly include systematic applications to sufficiently large patient samples, appropriately designed longitudinal studies, confirmation with histological results, and clinical trials.

A novel computerized tool to stratify risk in carotid atherosclerosis using kinematic features of the arterial wall

“…kNN is conceptually similar to PNN; kNN is a simple classifier, in which a feature vector is assigned the class that is most common among its k nearest neighbors, based on some similarity measure (distance function) [30]. The classification performance of kNN depends on the parameter k (in [3][4][5][6][7][8][9][10][11]) and the distance function (in {city block, Chebyshev, standardized Euclidean, cosine, Euclidean, Hamming, Jaccard, Minkowski}).…”

“…However, there is evidence that this marker is not sufficient [6], thereby motivating the development of computer-aided-diagnosis (CAD) systems, which incorporate additional markers to assist treatment planning. In particular, the development of CAD systems which are based on 2-D ultrasound (US) image analysis is considered as a grand challenge by the scientific community, because 1) traditionally, vascular physicians select US examination in diagnosis and follow-up for patients with CA, 2) the use of affordable imaging techniques, such as US, is crucial to avoid increasing the socioeconomic burden of the disease, and 3) 2-D US is traditionally available on commercial US systems [7], [8].…”

“…Regarding image-based features, plaque texture has been thoroughly investigated, with very encouraging results by CAD tools [7]. The morphology of the arterial wall has also been included in a limited number of CAD systems, without gaining as much attention [7].…”

“…The morphology of the arterial wall has also been included in a limited number of CAD systems, without gaining as much attention [7]. The role of kinematic features, which are estimated with motion analysis from B-mode US image sequences and represent dynamic phenomena and mechanical interactions occurring during the cardiac cycle (CC), remains unexplored, though.…”

A fully automated method for carotid plaques segmentation in ultrasound images based on motion estimation and level-set

“…The atherosclerotic process consists in the degeneration of the arterial walls and the formation of carotid plaques (CPs). The risk of cardiovascular events increases with the severity of carotid stenosis and also depends on CPs composition [1]. CA diagnosis, usually guided by US imaging, is based on the assessment of the stenosis degree due to the presence of CPs and on CPs composition study.…”