Magnetic Resonance Cholangiopancreatography: Image Quality Control And Clinical Application

Objective1. To evaluate the ability of Gadopentetate Dimeglumine (Gd-DTPA) enhancement to suppress high signals of the ventral small vessels on the background of pancreaticobiliary tree that infects the image qulity of magnetic resonance cholangiopancreatography (MRCP).2. To choose optimal concentration and content of Gd-DTPA solution as a oral gastrointestinal negative contrast agent for MRCP. To evaluate the role of Gd-DTPA solution to supress high signals from stomach and duodenum and to improve image qulity of MRCP.3. To combined with oral and Gd-DTPA enhancement to suppress noisy signals from stomach and duodenum and the ventral small vessel signals on the background of pancreaticobiliary tree to improve the image quality of MRCP.Materials and Methods1. 31 patients with suspected of biliary tract and pancreatic disease were chosen as object in this study. Gd-DTPA injection was provided by Guangzhou Consun Pharmaceutical Factory and its specification was 7.04g/15ml. All patients were performed before and after (1~5minutes, 5~10minutes and 10~15minutes) Gd-DTPA (adult: 0.2ml/kg) enhanced MRCP. Two experienced abdominal radiologists analyzed all MRCP images. Pre- and post- Gd-DTPA enhanced MRCPs of each patient were randomly presented to the observers. The window width and level of the pre- and postcontrast MRCPs of each patient were set at the same values. Observers were asked to determine the grade and score of 1~4 for intrahepatic duct branches,common hepatic duct, gallbladder, common bile duct and pancreatic duct. Discrepancies between observers were resolved by consensus. To analyze the signal-to-noise ratio (SNR) of liver, pancreas, common bile duct and pancreatic duct and the signal intensity ratio (SIR) of common bile duct relative to liver parenchyma and pancreatic duct to pancreatic parenchyma were calculated in both pre- and postcontrast MRCPs.2. In vitro experiment: Gd-DTPA injection 1, 2 and 3ml was separately added into warm water with the diluted solution being of 50, 100, 150, 200ml to be confected the different concentration and content of contrast administration. T1WI, T2WI, single thick-slice TSE sequence and HASTE were performed to these contrast agents and the signal intensity on each sequence was measured, enhancement ratio was calculated and to choose the optimal concentration of Gd-DTPA solution that were confected to different content as oral negative gastrointestinal contrast agent on MRCP. Clinical study: The optimal concentration and content Gd-DTPA solution regard as an oral negative gastrointestinal contrast agent of MRCP. Twenty-four patients were performed with MRCP before and after (5~10minutes and 10~15 minutes) oral negative gastrointestinal contrast agent. Observers were asked to determine the grade and score of 1-4 for intrahepatic duct, common hepatic duct, gallbladder, common bile duct and pancreatic duct. Discrepancies between observers were resolved by consensus. The method was same to part one.3. On the base of vitro experiment and clinical study, Gd-DTPA injection 2.0ml was added into 98ml water to mix 100ml that confected O.Olmol/L of contrast administration that was regarded as an optimal oral negative gastrointestinal contrast agent of MRCP. Twenty-one patients were performed with and wihtout oral combined with Gd-DTPA enhanced (10 minutes) MRCP. To analyze the show of common bile duct and pancreatic duct. The method of image quality analysis was same to the part 1 and part 2.4. Statistical analysis was porformed using SPSS 10.0 and using nonparametric Wilcoxon's test and independent-samples T test and X2 test. P>0.05 was regarded as no statistical difference; P<0.05 was regarded as statistical difference and P<0.01 was regarded as significant difference.Results1. After Gd-DTPA enhanced l~5minutes, 5~10 minutes and 10~15minutes, the show and average grade score of the pancreaticobiliary tree on MRCP had no significant difference (P>0.05). On Gd-DTPA enhanced MRCP images, the signal intensity of abdomen small vessels was decreased. On 2D FSE and 3D HEASTE MRCP images, the SNR of the common bile duct and pancreatic duct were no statistical difference (P=0.208, 0.094) between pre- and post Gd-DTPA enhancement, while the SNR of the liver parenchyma and the pancreas parenchyma on postcontrast MRCP images were significantly lower than those on precontrast MRCP images (p=0.000). The show of pancreaticobiliary tree structures after Gd-DTPA enhanced MRCP was improved than before enhancement due to suppress those signals from abdomen small vessles and the signal intensity of the liver and pancrease decreased so that the second branch of the intrahepatic bile duct, the common hepatic duct, the common bile duct and pancreatic duct of the pancreaticobiliary tree were clearer.2. In vitro experiment: tubes of warm water are hypointensity, the more the concentration of Gd-DTPA was, the more the signal intensity was increased on T1WI. When the concentration of Gd-DTPA was ^0.01mol/L, the contrast agent was hyperintensity on Tl WI. On T2WI, the more the concentration of Gd-DTPA was high, the more the T2 was shorted, and when the concentration was reached 0.015mol/L, the tube was as low as basic ground. On 2D TSE MRCP image, the contrals were hyperintensity and the concentrations from 0.0025mol/L to 0.03mol/L were hypointensity. On 3D HEAST MRCP image, the contrals were hyperintensity and when the concontrast of Gd-DTPA was ^ O.Olmol the contrast agents were hypointensity in which O.Olmol/L was lowest threshold concentration was chosen as MRCP oral negative gastrointestinal contrast agent. After the O.Olmol/L contrast agent confected into 50ml, 100ml, 150ml and 200ml were feeded to twenty patients, 50ml contrast agent incompelely suppressed the high signal from duodenum on MRCP images, while ^ 100ml contrast agents compelely suppressed these high signals. Therefore, we chose the Gd-DTPA solution in which the concentration was O.Olmol/L and the content was 100ml regarded as the MRCP oral negative gastrointestinal contrast agent (2ml was added into 98ml water to mix 100ml that confected O.Olmol/L of contrast administration).To more validate the results of vitro experiment, on MRCP images after using oral contrast agent with the concentration O.Olmol/L in 5~10minutes and 10~15minutes of 24 patients, the average grades scores of pancreaticobiliary tree had no significant difference. The scores of the third branch of the intrahepatic bile duct, the common bile duct and pancreatic duct between pre- and post- oral contrast agent was significantly difference (p<0.05). After oral contrast agent, the high signal from fluid of retention in the stomach and duodenum was suppressed completely. The image quality of MRCP after oral contrast agent was improved obviously than before oral contrast agent that the third branch of the intrahepatic bile duct, the common bile duct and pancreatic duct of the pancreaticobiliary tree was more clearly displayed.3. On oral combined with Gd-DTPA MRCP image of 21 patients, the signals from the superior mesenteric vein and its branches, the right and left portal veins and their branches and the high signal from fluid of retention in the stomach and duodenum were suppressed completely. The signal intensity of the liver and the pancreas parenchyma reduced obviously. Therefore, the image quality of MRCP after Gd-DTPA enhanced and oral contrast agent was improved obviously than those precontrase.The inner- and outer- hepatic biliary duct exhibit excellently on 3-dimensional thick-slice half-fourier acquisition single-shot fast spin echo (HASTE) MRCP images sequence, and the lumen is continual and the outline is direct and smooth and no obvious "ladder" sign, and pancreatic duct has no intermittent phenomenon or toothing sign. For the malignant obstruction of bile duct, the image is direct and vivid. Not only it can show the secondary change of the obstruction and also show the mass itself. Additionally, scan time of 3D HASET MRCP is very short and fit to the patients who are disoperative or can not bear the long time breathholding. However, the tiny filling defect is easily to be missed due to the cover of the fluid of the gallbladder and pancreas. 2D single slice TSE MRCP can vividly show the mini-biliary duct and pancreatic duct, it also help to show the small filling defect in the lumen. The gallbladder, the common biliary duct and pancreatic duct were clearly observed on the coronal section. The combination of left and right hepatic duct and the papilla were clearer on the inclined coronal section MRCP.Conclusion1. On Gd-DTPA MRCP images, the signal of overlapped vessels was suppressed which were from the superior mesenteric vein and its branches overlap with the pancreatic duct and the right and left portal veins overlap with the second branch of the intrahepatic bile duct, and the signal intensity of the liver and pancreas are decreased so that improve the contrast of the pancreaticobiliary tree, therefore MRCP image quality is dramatically improved.2. The O.Olmol/L contrast agent which is consisted of Gd-DTPA injection 2.0 ml and 98ml water might is optimal because its could suppressed compelely the high signal from the retention of the stomach and duodenum; moreover, the contrast agent formulate easily and has no additional cost and no side effects after oral so that the patients is prone to accept. It can regard as the oral MRCP contrast agent. After oral contrast administration in 5~10minutes, the high signal from gastrointestinal tract is suppressed compelely so that image quality of MRCP was improved obviously and the pancreaticobiliary tree was clearer.3. For patients suspected bile duct and pancreatic disease, oral combined with Gd-DTPA enhanced MRCP could completely suppress the high signal from the retention of the stomach and duodenum and from overlapped vessels of the left and right portal vein and the superior mesenteric vein and its branches. Therefore, the image quality of MRCP was more improved and united dynamic enhanced T1WI could contribute greatly to the diagnosis and critical diagnosis of the bile duct and pancreatic disease.4. On 3-dimensional thick-slice HASTE MRCP images, inner- and outer- hepatic biliary duct and pancreatic duct exhibited excellently and the lumen was continual and had no obvious "ladder" artifacts, and the malignant obstruction of bile duct was direct and clear. Additionally, scan time of the thick-slice MRCP was extremely short so that fit to the patients who are disoperative or could not bear the long time breathholding. However, the tiny filling defect is easily to be missed due to the cover of the fluid of the gallbladder and pancreas. 2D single slice TSE MRCP can vividly show the mini-biliary duct and pancreatic duct, it also help to show the small filling defect in the lumen. The gallbladder, the common biliary duct and pancreatic ductwas clear on the coronal section MRCP images. The combination of left and right hepatic duct and the papilla was clear on the inclined coronal section MRCP images. Combined the the 2D single slice TSE and 3D HASTE MRCP with coronal section and different inclined coronal section scan could provid more informations to the diagnosis and differentiated diagnosis of the bile duct and pancreatic disease.