Reliability of 3-D ultrasound measurements of cervical lymph node volume

Abstract

This study was undertaken to evaluate the reliability of three-dimensional (3-D) ultrasound in measuring cervical lymph node volume. Ultrasound examination of the neck was performed on 15 healthy subjects (eight men and seven women). For each subject, the volume of cervical lymph nodes was measured twice with 3-D ultrasound by two operators, to evaluate the reproducibility of measurements (interoperator variability). 3-D ultrasound measurements of cervical node volume were performed with and without using SonoCT® and XRES™ imaging. Each 3-D data set was reviewed and remeasured by the operators to evaluate the repeatability of measurements (intra-operator variability). Results showed that the reproducibility and repeatability of 3-D ultrasound volumetric measurements of cervical nodes was improved when SonoCT® and XRES™ imaging were used. There was a high repeatability of 3-D ultrasound measurements of cervical node volume (> 90%). A high reproducibility of measurements was found in the posterior triangle nodes (90.3% - 90.9%). When SonoCT® and XRES™ imaging were used with the scanning, there was a high reproducibility for parotid node measurements (87.4%) and a satisfactory level of reproducibility for submental (61.8%), submandibular (69.3%) and upper cervical node (79%) measurements. 3-D ultrasound is a useful and reliable method in measuring cervical lymph node volume. However, further studies to improve the reproducibility of 3-D ultrasound measurements of submental, submandibular and upper cervical node volume are suggested.

Introduction

Assessment of cervical lymph nodes is important to patients with head and neck cancers because it aids treatment planning and prediction of prognosis (van den Brekel et al. 1990; Ishii et al. 1991; Vassallo et al. 1993). Evaluation of posttreatment lymph nodes also helps assessment and monitoring of the treatment responses (Ahuja et al. 1996; van den Brekel et al. 1999). The size of lymph nodes is an important parameter in the assessment of lymph nodes. It is because pathologic lymph nodes tend to be enlarged. In patients with head and neck cancer, increase of nodal size in serial examinations indicates metastases of lymph nodes, whereas decrease in size of lymph nodes after treatment indicates positive response to the treatment (Ahuja et al. 1996; Ahuja and Ying 2003).

High-resolution ultrasound is a common imaging tool for the evaluation of cervical lymphadenopathy (Chang et al. 1990; van den Brekel et al. 1990; Baatenburg de Jong et al. 1991; Bruneton et al. 1994; Baatenburg de Jong et al. 1998). In routine ultrasound evaluation of cervical lymph nodes, nodal size is usually assessed by measuring the transverse diameters of the lymph nodes (Yuasa et al. 2000; Yusa et al. 2000; Ahuja and Ying 2002). However, it has been reported that the evaluation of nodal size by measuring the volume of lymph nodes is more accurate (Dietrich et al. 1997; Dietrich et al. 1999a; Dietrich et al. 1999b; Choi et al. 2001; Lewis et al. 2002). Nevertheless, previous studies assumed lymph nodes to be elliptical in shape and used the ellipsoid formula to calculate the nodal volume (Dietrich et al. 1997; Dietrich et al. 1999a; Dietrich et al. 1999b; Choi et al. 2001; Lewis et al. 2002). Three-dimensional (3-D) ultrasound has been reported to have a high accuracy in volumetric measurement of different structures both in vitro and in vivo (Riccabona et al. 1996; Chang et al. 1997; Chou et al. 1997; Kurjak et al. 2003; Ng et al. 2004; Pang et al. 2005). However, there is scant information about 3-D ultrasound volumetric measurements of cervical lymph nodes (Kim et al. 2002).

Before 3-D ultrasound is routinely used to measure the volume of cervical nodes, it is necessary to understand its reliability of measurements in cervical lymph nodes. Moreover, since 3-D ultrasound requires smooth sweep of scans for accurate measurements, lymph nodes in different regions of the neck have different adjacent structures, which may cause variation of the reliability of measurements. Therefore, this study was undertaken to investigate the reliability of sonographic volumetric measurements of cervical nodes in different regions of the neck.

Compound ultrasound imaging can provide images with a higher resolution when compared with conventional ultrasound (Entrekin et al. 2001; Kern et al. 2004). As high resolution images may allow more accurate outline of the nodal boundaries in 3-D ultrasound and provide more accurate results, the reliability of volumetric measurements of cervical nodes with and without using compound imaging was also compared in the present study.