HEPATOLOGY
limited in demonstration of partial thrombosis and occlusions of portal vein branches.
When inadequate Doppler tracings are obtained, this "technical failure" is often related to the presence of retrograde flow or to portal vein thrombosis and indicates the need for angiography (5). Use of color Doppler in the evaluation of portal flow is a very promising application for this new technology, particularly in differentiating artifactual, intraluminal echoes from soft, hypoechoic thrombi (3).
In current practice, real-time ultrasound, with the addition of Doppler, is a useful, noninvasive modality in the evaluation of portal vein patency. When the study is technically inadequate or results are equivocal, particularly in postsurgical cases, further testing is indicated, which at present relies on angiography as the "gold standard." Other noninvasive modalities currently undergoing evaluation include magnetic resonance imaging and computed tomography (6). These may be useful in identification of peripheral thrombi. The caveats that apply in the verification of results, optimization of technique and establishment of a gold standard apply equally. Futhermore, the ethics of performing invasive or operative tests solely for verification remains a major obstacle in the determination of the sensitivity and specificity of diagnostic techniques.