Preoperative evaluation of the entire hepatic vasculature in living liver donors with use of contrast-enhanced MR angiography and true fast imaging with steady-state precession

PURPOSE: To preoperatively assess the entire hepatic vasculature in living related liver donors with use of a combination of contrast material-enhanced magnetic resonance (MR) angiography and true fast imaging with steady-state precession (FISP). MATERIALS AND METHODS: Twenty-five living potential liver donors were examined preoperatively on a 1.5T Siemens Sonata system. Twenty-four underwent surgery and two had catheter angiography performed to delineate complex anatomy. Contiguous 5-mm-thick, sub-second true FISP images of the liver were initially obtained during breath-holding in axial and coronal planes (repetition time [TR]/echo time [TE], 3.2/1.6; flip angle, 70 degrees ). MR angiography was performed with use of a three-dimensional (3D) gradient-echo fast low-angle shot (FLASH) pulse sequence (TR/TE, 3.0/1.2; flip angle, 25 degrees ), with 40 mL of Gadolinium DTPA injected at a rate of 2 mL/sec. One precontrast and two postcontrast coronal 3D volumes were acquired, each in a 20-second breath-hold, and two subtracted 3D sets were calculated. Arterial anatomy was assessed with use of maximum-intensity projection, volume rendering, and multiplanar reformatting algorithms. Hepatic and portal venous anatomy was evaluated with use of the true FISP images and the venous phase of the MR angiogram. Visualization of hepatic arterial branches was noted. Visualization of portal vein branches was scored on a scale of 0-5. The presence of anatomic variants was noted. Vascular anatomy was confirmed at the time of surgery and at catheter angiography. RESULTS: Segmental branch vessels were visualized on MR angiography in the majority of cases. The segment four branch was identified in 96% patients. Variant arterial anatomy was seen in 50% of patients. MR angiography detected 10 of 11 arterial variants found at surgery and angiography. Visualization of portal vein branches was generally higher with true FISP compared to MR angiography. Twenty-four percent of patients had variant portal venous anatomy. Caudal hepatic veins were identified in 60% of patients, of which eight were significant (>5 mm). Hepatic and portal venous anatomy was accurately predicted by true FISP and MR angiography in all patients who went on to undergo surgery. CONCLUSION: Preoperative imaging with use of a combination of contrast-enhanced MR angiography and true FISP provides a comprehensive assessment of the entire hepatic vasculature in living liver donors.     

Detection of thrombosis in the portal venous system: comparison of contrast-enhanced MR angiography with intraarterial digital subtraction angiography

PURPOSE: To determine whether intraarterial digital subtraction angiography (DSA) can be replaced by contrast material-enhanced magnetic resonance (MR) angiography in the assessment of patency or thrombosis of the portal venous system in patients with portal hypertension. MATERIALS AND METHODS: Thirty-six patients with portal hypertension underwent contrast-enhanced MR angiography and intraarterial DSA for assessment of the portal venous system. The images were evaluated for vessel patency or thrombosis of the portal, splenic, or superior mesenteric vein. RESULTS: Of the 101 vessels evaluated, 42 were thrombosed. Overall sensitivity, specificity, and accuracy for the detection of thrombosis were 100%, 98%, and 99%, respectively, for MR angiography and 91%, 100%, and 96%, respectively, for DSA; differences between the imaging methods were not statistically significant. Only in four patients with six vessels (6%) were there discordant findings between MR angiography and DSA. CONCLUSION: Noninvasive contrast-enhanced MR angiography has the potential to replace intraarterial DSA as the standard method to assess the whole portal venous system.     

MR portal venography: preliminary results of fast acquisition without contrast material or breath holding

RATIONALE AND OBJECTIVES: The authors performed this study to evaluate the feasibility of using the steady-state free precession (SSFP) sequence to perform magnetic resonance (MR) venography of the portal venous system without the use of contrast material or breath holding. MATERIALS AND METHODS: Eleven patients underwent MR venography with the SSFP technique. Coronal three-dimensional images were obtained with respiratory triggering. Contrast material and respiratory suspension were not used. All patients had recently undergone at least one other imaging study (conventional angiography, transhepatic portal venography, ultrasound, or contrast-enhanced computed tomography), and these findings were correlated with those from MR venography. The structures evaluated were the main portal vein, right portal vein, left portal vein, superior mesenteric vein, and splenic vein. RESULTS: MR venography with SSFP accurately depicted the status of these veins in all cases except one. In this patient, MR venography depicted portal vein thrombus but could not indicate that it was tumor thrombus. CONCLUSION: MR venography with SSFP accurately depicted the portal venous system in 10 of 11 patients without the use of respiratory suspension or contrast material.     

MR imaging of the portal venous system: value of gradient-echo imaging as an adjunct to spin-echo imaging.

We evaluated the use of gradient-echo (GRE) as an adjunct to spin-echo (SE) MR imaging of the portal venous system. GRE imaging was performed in 31 subjects, 15 normal volunteers and 16 patients with documented portal venous disease (15 cases) or suspected disease (one case). Eight of 16 patients had venous thrombosis, five had focal thrombus, and three had complete occlusion. Six patients had extrinsic venous compression by tumor. Of the two other patients, one had an arteriovenous fistula and the other a falsely positive angiogram, suggesting portal vein occlusion. In normal subjects, GRE scans had excellent visualization of the portal venous system with high intravascular signal compared with surrounding tissues. Nine (60%) of 15 normal subjects and three patients had an artifact consisting of a curvilinear area of decreased signal that could mimic clot. In three of five patients with focal thrombus, clot was identified on GRE but not on SE images. In all three patients with occlusion, SE and GRE images demonstrated similar findings. In five of the six patients with extrinsic venous compression by tumor, SE and GRE studies showed similar findings. Of the two patients, an arteriovenous fistula was seen on GRE MR in one, and in the other, patency of the left portal vein was seen on SE and GRE images after angiography had suggested portal vein occlusion. Collateral vessels were seen in nine of 16 patients. In five of nine cases, GRE MR demonstrated more extensive collaterals than did SE MR. In summary, GRE MR provides a useful adjunct to standard SE MR imaging. Benefits include high contrast between vascular structures and surrounding tissues, reduced motion artifact, and rapid scanning within a breath-hold.     

Is true FISP imaging reliable in the evaluation of venous thrombosis?

OBJECTIVE: The objective of our study was to evaluate the accuracy of true fast imaging with steady-state precession (FISP) in the diagnosis of venous thrombosis using gadolinium-enhanced 3D T1-weighted gradient-echo images and correlative imaging as the gold standard. MATERIALS AND METHODS: Twenty-five MR examinations were retrospectively reviewed independently by two radiologists to rule out thrombosis in the central veins of the body. The presence of venous thrombus was assessed separately in 80 veins using true FISP and gadolinium-enhanced T1-weighted images. Diagnosis was confirmed by another imaging technique (sonography, CT, and/or conventional venography) in all positive cases. Negative examinations were confirmed using imaging, clinical follow-up, or both. RESULTS: Venous thrombosis was present in 25 veins in 18 patients. True FISP images had a lower sensitivity (66%) and specificity (70.9%) for the diagnosis of venous thrombosis than gadolinium-enhanced MR images (p < 0.01). CONCLUSION: True FISP images have lower sensitivity and specificity in the diagnosis of venous thrombosis than gadolinium-enhanced T1-weighted gradient-echo images. True FISP images should not be used exclusively for the diagnosis of venous thrombosis.     

The correlation between MR and angiography in portal hypertension

Forty-two MR examinations and hepatic panangiograms in 38 patients with portal hypertension were correlated with MR images to determine the ability of MR to detect portal vein hemodynamics. These studies were prospectively analyzed for degree of portal perfusion and direction of flow, portal vein thrombosis, and presence and type of shunt surgery. Thirty-three MR examinations were determined to have grade I (good) or II (fair) portal blood flow. Twenty-nine of these were grade I or II by angiography; the other four were grade IV. Of the eight cases documented as grade IV (hepatofugal portal blood flow) by angiography, none were considered grade IV by MR, suggesting that MR was unable to detect retrograde flow. The other case was not graded because of cavernous transformation of the portal vein. MR correlated well with angiography for the detection or absence of portal vein thrombus, agreeing with angiography in 41 of 42 cases. Two angiographically proven cases of portal vein thrombosis were correctly identified on MR. MR correctly identified the absence of portal vein clot in 39 of 40 angiographically negative cases. MR and angiography also agreed in 41 of 42 cases that a shunt was either present/absent or patent/occluded. The single error was due to inadequate MR scanning in the region of interest. The results show that MR cannot be used to grade blood flow in the portal vein. However, MR accurately detects portal vein thrombosis and the patency of surgical shunts.     

Coronary artery bypass graft patency: assessment with true ast imaging with steady-state precession versus gadolinium-enhanced MR angiography

PURPOSE: To compare the accuracy of multisection true fast imaging with steady-state precession (FISP) with gadolinium-enhanced magnetic resonance (MR) angiography for the detection of coronary artery bypass graft patency. MATERIALS AND METHODS: Twenty-five patients with coronary artery bypass grafts who had recently undergone conventional coronary angiography underwent MR angiography with a 1.5-T system. True FISP angiographic images were acquired in transverse and coronal planes. Coronal cardiac-gated MR angiography was performed with 0.2 mL per kilogram of body weight of gadopentetate dimeglumine injected at a rate of 2 mL/sec. With conventional angiography as the reference standard, the sensitivity, specificity, and accuracy of each technique for the detection of graft patency were determined. Image quality and duration of analysis were determined by two experienced radiologists. RESULTS: In 25 patients, 46 of 56 venous grafts were patent and 22 of 23 arterial grafts were patent. In all grafts at true FISP angiography, sensitivity for patency was 84% (57 of 68 grafts), specificity was 45% (five of 11 grafts), and accuracy was 78% (62 of 79 grafts). At MR angiography, sensitivity was 85% (58 of 68 grafts), specificity was 73% (eight of 11 grafts), and accuracy was 84% (66 of 79 grafts) (difference not significant). Image quality scores were similar with both techniques, but duration of analysis was significantly longer with MR angiography than with true FISP angiography (29 minutes 24 seconds vs 14 minutes 6 seconds, P <.001). CONCLUSION: Accuracy for detection of coronary artery bypass graft patency was similar with gadolinium-enhanced MR angiography and true FISP angiography, with a trend toward more false-positive findings for occlusion and reduced visualization of arterial grafts with true FISP angiography.     

MR imaging of portal venous thrombosis: correlation with CT and sonography

Fourteen patients with portal venous thrombosis (PVT) diagnosed by CT and/or sonography were studied with MR. Three of the 14 had portal hypertension. The MR findings were compared with those of eight patients with portal hypertension, but without CT or sonographic evidence of PVT. MR imaging showed portal venous thrombosis in all 14 PVT cases. Intraluminal thrombi of less than 5 weeks duration appeared markedly hyperintense relative to liver and muscle on both T1- and T2-weighted images. Older thrombi appeared hyperintense relative to liver and muscle in eight of 11 cases, but only on T2-weighted images. MR showed thrombi in 11% more portal vessels than did CT (MR = 30, CT = 27) and in 28% more vessels than did sonography (MR = 32, sonography = 25). MR also showed 24% more collateral vessels than did CT (MR = 31, CT = 25) and 50% more vessels than did sonography (MR = 33, sonography = 22). Third-echo images (echo time = 96 msec, repetition time = 1500-2150 msec) verified the presence of venous thrombi in 28 (93%) of 30 PVT vessels, and they differentiated flow-related intravascular signal from true thrombi in six (17%) of 36 portal hypertension vessels. We conclude that MR is a valuable tool for imaging portal vein thrombosis. MR is a good substitute for CT and can be more informative than sonography.     

MR evaluation of the portal vein in pediatric liver transplant candidates

Nine pediatric liver transplant candidates underwent preoperative MR evaluation of the portal vein and the inferior vena cava. Sonographic correlation was available in all patients and angiographic correlation was available in five. Pathologic correlation was obtained in seven cases either at liver transplantation or autopsy. MR demonstrated portal vein patency in three cases when it was not seen by angiography and confirmed portal vein patency in one patient when it was questionably identified on sonography. The portal vein was not seen on MR imaging in two cases when it was seen on sonography and angiography: in one case, it was small and to-and-fro flow was demonstrated angiographically; in the second case, the portal vein was occluded by tumor thrombus. Two vessels in two patients were misidentified by sonography and identified correctly by MR. These were an azygous continuation of the inferior vena cava and a large collateral vein in the portal region. Knowledge of the anatomy and documentation of vascular patency are essential in evaluation of patients before liver transplantation. In patients with complex anatomy or hemodynamics, it may be necessary to obtain this information from several imaging techniques (sonography, angiography, and MR).     

Portal hypertension evaluated by MR imaging

Thirteen patients with portal hypertension were examined by magnetic resonance (MR) imaging, using spin-echo sequences, and by visceral angiography. Data from this group were compared with the MR images and angiograms of 20 patients without portal hypertension. MR imaging demonstrated two of three cases of portal vein occlusion and four of five cases of occlusion of the retropancreatic portion of the superior mesenteric vein. Two thirds of the patients with portal hypertension and patent portal veins had marked MR signal within the main portal vein on MR images. Little or no signal was present in the portal veins of the 20 patients without portal hypertension. Our experience indicates that marked intraportal MR signal can be seen in patients with portal hypertension with or without venous occlusion. In some cases the size and distribution of venous collaterals allow one to distinguish between venous occlusion and other causes of portal hypertension. MR images confirmed the patency of distal splenorenal shunts in two patients studied.     

Magnetic resonance angiography with three-dimensional dynamic technique after contrast media administration for the study of the portal system ]

PURPOSE: To evaluate the feasibility of a contrast enhanced MR angiography (MRA) technique, using the latest 1.5 T MR tomoscan, to obtain optimal imaging of the portal system and compare the angiographic images with those obtained by color-Doppler and DSA. MATERIAL AND METHODS: Thirty patients (9 women and 21 men: average 53 years old) underwent contrast MRA of the portal vein, after portal hypertension had been diagnosed on the basis of clinical and chemical data and by color-Doppler. We used a dynamic 3D FFE T1-weighted breath - hold sequence during the arterial and venous phase after administering. 0.2 mmol/Kg of gadolinium-DTPA were at the rate of 2 ml/s. The contrast bolus was monitored using a 2D FFE T1-weighted sequence on a coronal plane. A FFE T1-weighted sequence was performed on axial plane before and after the dynamic sequence to obtain evaluate the a hepatic parenchyma. In the post processing phase MIP (maximum intensity projection) were reconstructed. We considered the patency of the portal venous system and the presence of cavernomatous and collateral circles; portal thrombosis was classified as partial or complete and as proximal or distal. RESULTS: Good quality MR angiographic images were obtained in 28 of the 30 cases examined; in 2 patients movement artefacts compromised the image quality. We observed a concordance between MRA and Doppler ultrasound in 79 vessels out of 84 (94%). A 97.5% concordance was found between MRA and DSA (82 vessels out of 84) with a sensitivity of 100% and a specificity of 97.3%. MRA was superior to DSA and Doppler ultrasound for evaluating large collateral shunts, above all gastro-esophageal and paraumbilical shunts, and complex anatomical conditions. CONCLUSIONS: Where available, advanced MRA technology with contrast enhancement should be used as a routine modality to study the anatomy and pathology and the portal system in all patients in whom Doppler ultrasound has yielded doubtful information. MRA is well-suited to obtain good vascular imaging before surgical or interventional procedures.     

MR angiography and dynamic flow evaluation of the portal venous system

We studied the value of MR angiographic techniques in imaging the portal venous system. Projection angiograms were created by postprocessing a series of two-dimensional, flow-compensated gradient-echo images. Flow velocity was determined by a bolus-tracking method with radiofrequency tagging and multiple data readout periods. Each image was acquired during a breath-hold. MR angiography was applied to six normal subjects and four patients with abnormal hemodynamics in the portal venous system. Flow velocity determined by MR was correlated with the results of duplex sonography. The main portal vein and intrahepatic branches were shown in all cases. Portosystemic collaterals were identified in all patients with portal hypertension. In normal subjects, peak flow velocities (17.9 +/- 2.8 cm/sec) on MR correlated well with values determined by duplex sonography (17.5 +/- 2.2 cm/sec) (r = .846, p less than .04). Reversed portal blood flow was shown in two patients. One patient with portal vein thrombosis had no evidence of flow by MR angiography. Our results indicate that MR angiography can provide a three-dimensional display of normal and abnormal vascular anatomy as well as functional information in the portal venous system.     

Diagnosis of portal vein thrombosis: value of color Doppler imaging.

This study was undertaken to determine the accuracy of color Doppler imaging in the diagnosis of portal vein thrombosis. Two hundred fifteen patients were studied with color Doppler imaging to determine patency of the main portal vein. Sonographic findings were confirmed in 75 patients, aged 19 to 66 years. Correlation with angiography was obtained in 13 patients, and surgical correlation was obtained in the remaining 62. Nine patients had portal vein thrombosis on the basis of these gold standards. Sonograms were classified as showing either patency or thrombosis, depending on the ability to show color flow within the main portal vein. Agreement between sonography and angiography or surgery was found in 69 patients (61 patent, eight thrombosed). One patient with a patent portal vein at sonography was found to have a thrombosed vessel at surgery, whereas five patients without portal venous flow at sonography had patent vessels at angiography (one patient) or surgery (four patients). Overall sensitivity and specificity for detection of portal vein thrombosis were 89% and 92%, with an accuracy of 92%, a false-negative rate of 0.11, a negative predictive value of 0.98, and a positive predictive value of 0.62. We postulate that the majority of errors in our study occurred in vessels that, although patent, had only sluggish flow, which could not be resolved because of technical limitations. We conclude that color Doppler imaging is a valuable screening procedure for the assessment of portal vein patency. If the sonogram shows a patent portal vein, no further studies are required. However, a lack of demonstrable flow does not always indicate thrombosis, and other imaging studies should be performed for confirmation.     

动态增强磁共振门静脉造影在门脉高压症诊断中的价值

目的:探讨动态增强磁共振门静脉造影(DCE MRP)在门脉高压症诊断中的价值。方法:对门脉高压症组23例和正常对照组15例行DCE MRP检查。观察23例门静脉高压症在DCE MRP上的表现。结果:对照组15例均清楚显示门静脉。门脉高压症表现为门静脉增粗(21例)、脾静脉迂曲扩张(23例);门静脉分支级数减少(18例);门静脉延迟显影(7例);显示侧枝循环静脉(6例);门静脉血栓形成(3例)。结论:DCE MRP是评价门静脉的一项快速无创的技术,可准确显示门静脉高压症门静脉系统的病理改变。     

Contrast-enhanced three-dimensional MR portography.

Three-dimensional (3D) magnetic resonance (MR) portography with contrast material enhancement is a fast means of evaluating the portal venous system that has some advantages over currently used modalities, such as digital subtraction angiography, helical computed tomography, ultrasonography, and nonenhanced MR angiography with time-of-flight and phase-contrast techniques. With contrast-enhanced 3D MR portography, a first-pass study of the mesenteric vasculature is performed after rapid bolus injection of gadopentetate dimeglumine; a 3D fast field echo sequence is used, which can demonstrate the intrahepatic and extrahepatic portal venous system clearly. Repeated sequences after administration of gadopentetate dimeglumine allow separate demonstration of the splanchnic arteries and portomesenteric veins. The images are reconstructed by means of maximum-intensity projection postprocessing, and a subtraction technique can be used to eliminate arterial enhancement and demonstrate portosystemic shunts. The coronal source images simultaneously demonstrate parenchymal lesions of the liver, pancreas, biliary tract, and spleen. This technique is clinically indicated in portosystemic shunt, portal vein thrombosis, hepatocellular carcinoma, pancreatobiliary tumor, hepatic vein obstruction, differentiation of splanchnic arterial from portal venous disease, and gastrointestinal hemorrhage. Its limitations include allergic reactions to contrast media, inappropriate positioning of the 3D acquisition slab, respiratory motion artifacts, and pseudodissection.     

Pancreatic portal cavernoma in patients with cavernous transformation of the portal vein: MR findings

The purpose of the article was to prospectively evaluate the MR findings of pancreatic portal cavernoma in a consecutive series of patients with cavernous transformation of the portal vein. This study was approved by the review board of our institution, and informed consent was obtained. The clinical and biological data and the MR imaging for 20 patients (11 female, 9 male; median age, 49 years) with cavernous transformation of the portal vein and no evidence of previous pancreatic disease were reviewed. The presence of pancreatic portal cavernoma (defined as intra- and/or peripancreatic portal cavernoma), morphological changes in the pancreas, biliary and ductal pancreatic abnormalities, and extension of the portal venous thrombosis were qualitatively assessed. Fifteen patients (75%) had pancreatic portal cavernoma with collateral formation in the pancreas and/or collaterals around the pancreas seen on dynamic contrast-enhanced MR sequences: three patients had both intra- and peripancreatic portal cavernoma, six had intrapancreatic portal cavernoma alone and six had peripancreatic portal cavernoma only. The presence of intra- or peripancreatic portal cavernoma was significantly associated with extension of the thrombosis to the splenic and superior mesenteric veins (p = 0.05). Morphological changes in the pancreas, heterogeneity on T2-weighted sequences and main ductal pancreatic abnormalities were seen in two, four and two patients, respectively. All these patients had intrapancreatic portal cavernoma. Bile duct dilatation was observed in 13 (65%) patients: among them three had extrahepatic dilatation only and these three patients had associated intrapancreatic portal cavernoma. In patients with cavernous transformation of the portal vein, intra- or peripancreatic portal cavernoma is common. In conclusion, intra- or peripancreatic portal cavernoma was only observed in patients with extension of the thrombosis to the splenic vein and/or the superior mesenteric vein.     

3D CE-MRA在评价腹部静脉系统中的应用

目的探讨三维增强磁共振血管成像（3DCE—MRA）技术在腹部静脉系统中的应用价值。方法收集具有完整资料的16例病例，先行常规MRI检查，随后经静脉注射Gd—DTPA20—30mL后连续三次行3DCE—MRA采集数据，对静脉系统原始图像进行MIP重建。主要观察指标为下腔静脉、脾静脉、肝静脉、肠系膜上静脉、门静脉形态变化。结果16例患者中，门腔静脉正常者9例，3DCE—MRA清晰显示下腔静脉、肝静脉、脾静脉、肠系膜上静脉、门静脉结构；门静脉高压3例，示脾静脉增宽迂曲，门静脉主干增宽及明显侧支循环形成；2例布加氏综合征；1例下腔静脉血栓形成；1例门静脉海绵样改变。结论3DCE—MRA是一种安全、敏感性高的血管造影技术，可提高腹部静脉系统疾病的术前诊断。     

门静脉瘤及少见并发症的3DDCE-MRA诊断

目的：回顾性分析门静脉瘤（portal vein ametryem,PVA）合并少见并发症的影像学特点,并探讨三维增强磁共振血管成像（3D DCE-MRA）对PVA的诊断价值。方法：利用3D DCE-MRA对10例PVA患者进行回顾性分析,显示及记录PVA的部位、大小、范围及并发症的情况。结果：10例PVA患者,瘤体位于脾静脉3例,肝内门静脉3例,肠系膜上静脉与脾静脉汇合部2例,门静脉主干2例。PVA最大径从8.39cm到2.00cm不等。PVA患者伴发多种少见并发症,通过3D DCE-MRA不同的三维重组技术能很好的显示。结论：3D DCE-MRA有效的诊断PVA,充分的显示门静脉的并发症,为临床治疗提供更多的影像学资料。     

Angiography of liver transplantation patients

Over 45 months, 119 angiographic examinations were performed in 95 patients prior to liver transplantation, and 53 examinations in 44 patients after transplantation. Transplantation feasibility was influenced by patency of the portal vein and inferior vena cava. Selective arterial portography, wedged hepatic venography, and transhepatic portography were used to assess the portal vein if sonography or computed tomography was inconclusive. Major indications for angiography after transplantation included early liver failure, sepsis, unexplained elevation of liver enzyme levels, and delayed bile leakage, all of which may be due to hepatic artery thrombosis. Other indications included gastrointestinal tract bleeding, hemobilia, and evaluation of portal vein patency in patients with chronic rejection who were being considered for retransplantation. Normal radiographic features of hepatic artery and portal vein reconstruction are demonstrated. Complications diagnosed using results of angiography included hepatic artery or portal vein stenoses and thromboses and pancreaticoduodenal aneurysms. Intrahepatic arterial narrowing, attenuation, slow flow, and poor filling were seen in five patients with rejection.     

MR imaging of portal vein thrombosis

MR imaging is emerging as a potential means of detecting portal venous thrombosis (PVT). Therefore, we attempted to establish specific criteria with which to diagnose PVT on conventional spin-echo images. In a retrospective review of 342 consecutive abdominal MR scans performed with a 0.5-T magnet, we identified nine patients with persistent signal in the portal vein and used the findings in these patients to establish criteria with which to diagnose PVT. We subsequently applied these criteria to 109 additional consecutive abdominal MR scans performed with the same magnet. Fifteen cases were found in which all images showed either (1) signal involving the entire width of the portal vein lumen, which approximated (with T1 weighting) and exceeded (with T2 weighting) the intensity of the hepatic parenchyma in images in which the hepatic veins showed a complete flow void or (2) complete nonvisualization of the portal vein and its major branches in images that showed a flow void in portal venous collaterals and hepatic veins. All patients had unequivocal findings of PVT on at least one other imaging study (CT or sonography) or at surgery. Although the sensitivity of these signs could not be calculated, their specificity was 100%. We conclude that in the presence of these signs, the diagnosis of PVT can be made with confidence.