Segmentation of the luminal border in intravascular ultrasound B-mode images using a probabilistic approach

Abstract

Intravascular ultrasound (IVUS) is a catheter-based medical imaging technique that produces cross-sectional images of blood vessels and is particularly useful for studying atherosclerosis. In this paper, we present a computational method for the delineation of the luminal border in IVUS B-mode images. The method is based in the minimization of a probabilistic cost function (that deforms a parametric curve) which defines a probability field that is regularized with respect to the given likelihoods of the pixels belonging to blood and non-blood. These likelihoods are obtained by a Support Vector Machine classifier trained using samples of the lumen and non-lumen regions provided by the user in the first frame of the sequence to be segmented. In addition, an optimization strategy is introduced in which the direction of the steepest descent and Broyden-Fletcher-Goldfarb-Shanno optimization methods are linearly combined to improve convergence. Our proposed method (MRK) is capable of segmenting IVUS B-mode images from different systems and transducer frequencies without the need of any parameter tuning, and it is robust with respect to changes of the B-mode reconstruction parameters which are subjectively adjusted by the interventionist. We validated the proposed method on six 20 MHz and six 40 MHz IVUS stationary sequences corresponding to regions with different degrees of stenosis, and evaluated its performance by comparing the * Corresponding author segmentation results with manual segmentation by two observers. Furthermore, we compared our method with the segmentation results on the same sequences as provided by the authors of three other segmentation methods available in the literature. The performance of all methods was quantified using Dice and Jaccard similarity indexes, Hausdorff distance, linear regression and Bland-Altman analysis. The results indicate the advantages of our method for the segmentation of the lumen contour.

Computerized segmentation of pulmonary nodules depicted in CT examinations using freehand sketches

“…This method is capable of depicting the discrepancy (agreement) between two 2D/3D outlines and has been used recently by a number of investigations. [38][39][40][41][42] Although the Bland-Altman plot 43 is widely used to analyze the agreement between two measures, it is not suitable for this study. The Bland-Altman is designed to visualize the relationship between two one-dimensional measures (assays), where the mean of the two measurements is used as the abscissa (xaxis) value, and the difference between the same two measurements is used as the ordinate (y-axis) value.…”

A region‐based segmentation method for ultrasound images in HIFU therapy

“…To date, many methods have been developed for the segmentation of various ultrasound images, including breast ultrasound images, 3,4 liver tumor ultrasound images, 5 echocardiography, 6,7 prostate ultrasound images, 8,9 intravascular ultrasound images, [10][11][12][13] and some ultrasound images of other organs or tissues. 14 As Huang et al 15 indicated, the active contour model (ACM) was widely used in the segmentation of all types of ultrasound tumor images in past decades.…”

A novel ultrasound based approach for lesion segmentation and its applications in gynecological laparoscopic surgery

“…In addition, the compute-aided segmentation is also used in our diagnostic tool. It applied the image processing technology to analyze the medical image and facilitates the location of the lesion, which has been successfully applied in many medical fields, such as detection of atherosclerotic plaque [3], retinal hemorrhage [4] and femoral head necrosis [5].And whether the manual diagnosis or the compute-aided segmentation relies on the acoustic characteristics because it facilitates the discrimination between different tissues and lesions in vivo.…”