Experimental pulmonary emboli detected using magnetic resonance.

Experimental pulmonary emboli that were labelled with non-magnetic barium threads and produced using aminocaproic acid were introduced into the internal jugular veins of five dogs. Transverse axial magnetic resonance (MR) images (0.35 T, 15 MHz) and CT scans (2.0 or 4.8 seconds, G.E. 9800 or 8800 scanner) were obtained from the lung apex to base in each animal. The MR images were gated to the cardiac cycle, and spin echo techniques were used (TR = gated to the cardiac cycle; TE = 28 and 56 msec). Nineteen sites of embolism were determined from the CT scans. Two observers, who had no knowledge of the number or position of the emboli, individually assessed the MR images and marked the sites of emboli on clear acetate overlays. Each observer detected 12 of 19 emboli (63%) and each had one false positive result. Of the 19 emboli, six were central or parahilar and 13 were in the outer two thirds of the lungs. Three (first observer) or four (second observer) of the seven false negative results were central. In retrospect, two central emboli were mistaken for hilar fat. One peripheral embolus was not visible, even in retrospect. The potential for MR to demonstrate relatively small pulmonary emboli has been shown. Clinical trials in patients seem warranted.