[False image of portal thrombosis in MRI: pitfalls in portal hypertension. Apropos of Budd-Chiari syndrome].

This observation illustrates how difficult it is to interpret the intravascular signal observed with MRI in cases of portal hypertension. A signal occupying the entire lumen of the portal trunk and branches can be seen, which is constantly observed in the axial as well as coronal planes in the 3 series studied. It is hypointense relative to the liver on T1-weighted images, hyperintense on T2-weighted images: this appearance therefore perfectly mimicks portal thrombosis. Ultrasound combined with pulsed Doppler demonstrates the stagnant blood within the patent portal system. This kind of artifact, which is produced by a very slow flow or even by the standstill of flow, is much more difficult to recognize than the common flow artifacts; the criteria put forward by various authors to distinguish between artifacts and thrombosis are of no avail in this case. Examinations must able to confirm or rule out portal thrombosis in the case of portal hypertension with deceleration of flow. In practice, ultrasound studies, combined with pulsed Doppler in the best cases, is therefore indicated as a first-intension technique.