Respiratory Impedance in a Mouse Model of Asthma using Hyperpolarized He MR Imaging

Introduction: Asthma is characterized by inflammation and constriction of the upper airways (airway hyper-responsiveness), when exposed to ambient irritants. There is currently substantial interest in understanding asthma pathophysiology through the use of animal models [1]. In these studies, animals are typically treated with the bronchoconstrictor Methacholine (MCh), and airway hyperresponsiveness is then monitored in vivo by respiratory impedance (Flexivent) [2], which monitors increases in airway smooth muscles constriction of the airways. Although Flexivent provides reliable and robust estimates of the Newtonian resistance (Rn), elastance (G) and compliance (H) of the lung, it provides only a global measurement of impedance. We recently demonstrated that it is possible to visualize in vivo the time course of MCh induced broncho-constriction in mice using hyperpolarized (HP) He gas MR imaging [3]. In this work, we observe dynamic changes in airway size using 2D HP He MR images obtained at multiple time points after the MCh injection. From the measured diameters of the visible airways, we then calculate the airway contribution to the global impedance. Thus, HP He MR imaging provides a novel method to precisely assess regional, time dependent estimates of Newtonian resistance in mouse models of asthma. Methods: He was polarized using optically pumped Rb vapor in a commercial polarizer (MITI, Durham NC). Male BalbC mice were mechanically ventilated [4] with a 0.25 ml tidal volume of HP He and O2 according to a Duke-approved IACUC protocol. Ten 2D, HP He radial images were acquired every 12 s at peak inspiration (matrix=128x128, FOV=20mm, BW=32.5 kHz, number of projections=400, TR/TE=5.0/0.272 ms) by applying 20, variable flip angle RF pulses [5] per breath. MCh was administered after the first 2D image acquisition. Images were processed in MatlabTM using a Sobel edge detection filter to enhance the airways edges. Five points were picked along each edge of 9-11 airways (3 generations). A linear least squares fitting was performed for the points picked along the trachea and a third order polynomial non-linear least squares fit was performed for the points along the other airways. Distance between the resulting curves was used to obtain the airway diameters. The Rn for air was then calculated using 4 8