A Semi-Automated Technique for Internal Jugular Vein Segmentation in Ultrasound Images Using Active Contours

Abstract

The assessment of the blood volume is crucial for the management of many acute and chronic diseases. Recent studies have shown that circulating blood volume correlates with the cross-sectional area (CSA) of the internal jugular vein (IJV) estimated from ultrasound imagery. In this paper, a semi-automatic segmentation algorithm is proposed using a combination of region growing and active contour techniques to provide a fast and accurate segmentation of IJV ultrasound videos. The algorithm is applied to track and segment the IJV across a range of image qualities, shapes, and temporal variation. The experimental results show that the algorithm performs well compared to expert manual segmentation and outperforms several published algorithms incorporating speckle tracking.