MR imaging parameters in the study of lung water. A preliminary study.

The use of magnetic resonance (MR) to evaluate lung water is made difficult by several factors: paucity of proton signal from normal lung, respiratory and cardiac motion, and long relaxation times of lung fluids. To optimize scanning parameters for this use, and to test MR's ability to detect and quantitate regional and temporal variations in signal intensity in hydrostatic pulmonary edema, in vivo experiments were performed with a 0.5 tesla whole body MR imaging device. Human volunteers were studied in prone and supine positions using spin echo technique (TE = 30 msec) with varying TR, and with respiratory and cardiac gating. In addition, sedated, intubated, chronically prepared sheep were paralyzed to control extraneous motion and allow the use of a high frequency ventilator, thereby eliminating respiratory gating. Elevated pulmonary hydrostatic pressure was induced in these sheep by inflation of a left atrial balloon. Relative signal intensity from the lung rises with lengthening TR. Cardiac gating diminishes motion artifact, but masks extravascular water by enhancing signal from slowly flowing blood by an average of 44%. A gravity-dependent gradient of signal intensity predictably shifts in supine and prone positions. The use of longer TRs, respiratory gating, and cardiac gating all proportionally prolong data acquisition times to an objectionable degree. Without the use of gating, a gradual rise in relative signal intensity is seen in the sheep lung following the establishment of elevated hydrostatic pressure in the pulmonary circuit, and is most pronounced in the dependent portion of the lung.(ABSTRACT TRUNCATED AT 250 WORDS)