ABSTRACT
The efficacy of non-breath-hold magnetic resonance (MR) cholangiography at mid-field strength (0.5 Tesla) was evaluated for delineating biliary anatomy and the cause and extent of biliary obstruction. We performed 65 MR cholangiograms on a mid-field 0.5 Tesla MR unit and correlated them with contrast cholangiography and/or surgery. MR cholangiography was found to be both sensitive and specific in the detection of biliary obstruction and in the definition of its cause (sensitivity 98%, specificity 100%, positive predictive value 100%, negative predictive value 85.7%, accuracy 98%). MR cholangiography accurately predicted the level of obstruction in 94 per cent of strictures. Normal caliber intra-hepatic biliary radicles were visualised in only 6 per cent of the MR cholangiograms. In contrast, 94 per cent of dilated intrahepatic biliary radicles were demonstrated. The confluence, and right and left hepatic ducts were visualized in 98 per cent; the gall bladder in 65 per cent; the cystic duct in 45 per cent and the cystic duct insertion in 25 per cent. The extrahepatic bile duct was seen in 82.7 per cent. A normal caliber pancreatic duct was seen in 18 per cent while a dilated pancreatic duct was seen in 86 per cent. The pancreatico-biliary junction was visualised in 7 per cent. Non-breath-hold MR cholangiography at midfield strength is a highly accurate method of evaluating the cause and level of biliary obstruction, comparable to high-field MR cholangiography. The spatial resolution however is inadequate for the detection of variations in biliary or pancreatic ductal anatomy when the ducts are of normal caliber.