CLOT-BLOOD CONTRAST IN FAST GRADIENT-ECHO MAGNETIC-RESONANCE-IMAGING

RATIONALE AND OBJECTIVES. Contrast between clot and blood in magnetic resonance imaging (MRI) at 1.5T using fast gradient-echo pulse sequenc es (fast GRE), with 8 ms < TR < 20 mseconds was studied both in vitro and in clinical human deep venous thrombosis (DVT) to assess whether g ood contrast could be obtained at such short repetition times and at c linically relevant flow rates. METHODS. In vitro studies used an appar atus that contained flowing MnCl2[aq] (water adjusted with manganese c hloride to have T1, T2 similar to blood) and an immobilized clot (T1, T2 similar to those in DVT) for flow velocities between 0 and 16.5 cm/ sec. Seven patients with DVT were imaged with the fast GRE sequences t o observe the clot-blood contrast in vivo. RESULTS. Peak contrast-to-n oise ratio (CNR) was achieved using flip angles between 20-degrees and 40-degrees (increasing with flow velocity) with or without radiofrequ ency ''spoiling,'' consistent with a natural spoiling effect of flow. The CNR between MnCl2[aq] and clot decreased less than 10% as TR was r educed 56% from 18 mseconds to 8 mseconds (30-degrees flip angle). In four patients with nonocclusive DVT, fast GRE imaging provided good co ntrast while in occlusive cases (three patients) the contrast was not as good as conventional GRE sequences with longer TR values (TR = 33 m seconds). CONCLUSION. A fast GRE sequence with TR = 8 mseconds, TE = 3 mseconds, and a flip angle = 40-degrees is a promising approach to sp eeding up the detection of nonocclusive clinical DVT.