Quantitation of lung water by NMR imaging: A preliminary study

This study investigates the relation of the NMR spin-lattice relaxation time (T/sub 1/) to tissue wetness in an animal model of pulmonary edema. Imaging was performed with a whole-body resistive magnet prototype NMR imager (Technicare, Inc.) operating at 0.15T. Following control NMR imaging, mongrel, mongrel dogs (n=4) were anesthetized and subjected to saline pulmonary lavage (15 ml/kg) while in a lateral position, repeated 2-5 times to produce mild to severe edema and hypoxemia (prelavage, on 100%, P/sub a/O/sub 2/ > 300 mm Hg; post lavage P/sub a/O/sub 2/ 65-85 mm Hg). The animals were maintained on positive pressure ventilation with halothane/O/sub 2/. Model P/sub a/O/sub 2/ stability was 15%/hr. Axial NMR images were obtained at the level of the carina. Partial saturation pulse sequences (900-1800-spin echo) were used in which the T/sub 2/ parameter was held constant (time to echo, T/sub E/) at 30 msec, while the T/sub 1/ parameter (repetition rate, T/sub R/) varied in multiple steps from 500-4500 msec, allowing production of a parametric image of T/sub 1/ distribution. Following sacrifice, two 1-cm thick transverse sections were removed from the lungs at the level of the carina. One set of transverse sections was analyzed intact and from the adjacent section four samples (approx. 5 cm/sup 2/) were excised. Gravimetric measurement of dry and wet weight was done, and comparison made to T/sub 1/ measurements found by in vivo imaging. Tissue wetness bore a linear relation to T/sub 1/ both when lung samples (n=12, r=0.7, rho<0.02) and whole slices (n=8, r=0.66, rho≅ 0.05) were used. The data also corresponds well to previous data determined in vitro at a different field strength, using a different animal model. Thus, T/sub 1/ calculation may allow quantitation of lung water