Quantitative analysis of the velocity and synchronicity of diaphragmatic motion: dynamic MRI in different postures

Abstract

The objectives of this study were to assess the relationship between right and left hemidiaphragmatic motions during breathing in normal subjects and to investigate alterations in lung motion with changes in posture, using dynamic magnetic resonance (MR) imaging. Imaging was conducted with a 1.5-T MR scanner using fast imaging employing steady-state acquisition with a torso coil. Eight healthy subjects were instructed to breathe from end-inspiration to end-expiration as slowly and as deeply as possible. Imaging and breathing were started together to afford sequential images on the coronal plane. Imaging sequences were performed in supine, prone, left lateral decubitus and right lateral decubitus postures. The component of movement of the most cephalic point in the cephalocaudal axis was measured, and the diaphragmatic excursion (maximum hemidiaphragmatic displacement), synchrony and velocity of the right and left hemidiaphragmatic motions were calculated during the expiratory phase and the inspiratory phase, respectively. Excursion was greater in the right hemidiaphragm in most postures, except the left lateral decubitus. In supine and prone postures, both hemidiaphragms moved synchronously in both inspiratory and expiratory phases. In both lateral decubitus postures, the hemidiaphragms moved asynchronously with different velocities in the expiratory phase but with the same velocities in the inspiratory phase. The method described here allowed the assessment of diaphragmatic motions. Motions in the right and left hemidiaphragms changed with posture. In addition, diaphragmatic motion differed between expiratory and inspiratory phases. This study suggests the further potential of dynamic MR imaging for the evaluation of pulmonary functions or deficiencies.

Introduction

Lung motion during respiration has been investigated using several imaging modalities. Investigators initially used fluoroscopy [1], [2], [3] and ultrasound [4], [5], [6]. In most of these studies, the displacements and excursions of the diaphragm or the chest wall were measured on images obtained at two different phases of respiration: end-inspiration and end-expiration. In normal subjects, Wade [1] and Alexander [2] reported that diaphragmatic excursion was greater on the left side, whereas Simon et al. [3] and Houston et al. [5] reported that it was greater on the right side.

Recently, magnetic resonance imaging (MRI) has been used for the assessment of lung motion because of its ability to provide good tissue contrast [7], [8], [9], [10]. MRI is not affected by distortion due to magnification or parallax characteristic of fluoroscopic techniques, thus providing more precise and reproducible measurements in different subjects or examinations [9]. Furthermore, MRI allows one to perform a dynamic study during breathing, where a series of images was obtained at high temporal resolution during respiration [9], [11], [12], [13]. These previous studies suggest that MRI is more attractive than X-ray computed tomography due to its noninvasive nature and its usefulness for the assessment of abnormal lung motions accompanying several diseases such as diaphragmatic paralysis and chronic obstructive pulmonary disease [11], [12], [13], [14], [15]. In most of those studies, however, only total excursion was quantified, and they reported conflicting conclusions regarding diaphragmatic differences in hemidiaphragmatic excursion on the right and left sides. Although dynamic MRI was qualitative, Gierada et al. [9] described that it could reveal minor asynchrony in right and left hemidiaphragmatic motions. This has highlighted the need to establish a quantitative method for the assessment of a more detailed dynamic diaphragmatic motion (e.g., synchrony between right and left, or the velocity of diaphragmatic motion in dynamic MRI).

In previous reports that described dynamic MRI, each diaphragm level was measured during breathing, and time–diaphragmatic length curves were generated [12], [16]. However, the curves were used mainly to detect maximum expiration/inspiration and to compare the highs of right and left diaphragms at each point; therefore, diaphragmatic motion was analyzed statically. To evaluate detailed diaphragmatic motion, we should deal with it as a continuous dynamic motion, and a new analysis is necessary to perform this evaluation.

Pulmonary ventilation is heterogeneous and location-dependent in the lungs [17], [18], [19], [20], being greater in the dependent lung where gravity is thought to play an important role [21], [22], [23]. However, there have been few studies of lung motion in different postures. Froese and Bryan [24], using fluoroscopy, reported diaphragmatic displacements in the supine and left lateral decubitus postures, but there have been no studies assessing posture-dependent lung motion dynamically.

Here we propose new quantitative analysis methods to assess dynamic diaphragmatic motion in dynamic MRI studies. Using such analyses, we quantified the total excursion, right–left synchrony and velocity of the right and left hemidiaphragmatic motions on the coronal plane during breathing. Furthermore, we investigated whether those parameters were altered by different postures: supine, prone, right lateral decubitus and left lateral decubitus.