AJR
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Carlos, R. C.
Right arrow Articles by Fendrick, A. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Carlos, R. C.
Right arrow Articles by Fendrick, A. M.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
Hotlight (NEW!)
Right arrow
What's Hotlight?

Developing a Prediction Rule to Assess Hepatic Malignancy in Patients with Cirrhosis

Ruth C. Carlos1, H. Myra Kim2, Hero K. Hussain1, Issac R. Francis1, Hanh V. Nghiem1 and A. Mark Fendrick3

1 Department of Radiology, MRI Section, University of Michigan, 1500 E. Medical Center Dr., UH B2B311, Ann Arbor, MI 48109-0030.
2 Center for Statistical Consulting and Research, University of Michigan, Ann Arbor, MI 48109-0030.
3 Department of Medicine, University of Michigan, Ann Arbor, MI 48109-0030.



View larger version (10K):

[in a new window]
  		Fig. 1. —Graph of receiver operating characteristic curve shows performance of regression model for detecting hepatic malignancy in patients with cirrhosis. Regression model used arterial enhancement, venous or delayed phase washout, lesion size, number of lesions, and {alpha}-fetoprotein level as predictors and considered dysplastic nodules as malignancy. Area under curve is 0.82, suggesting very good discriminant ability of predictive model in correctly categorizing malignant lesions. Diagonal line represents test of no discriminative ability or test that is not better than chance.

 


View larger version (128K):

[in a new window]
  		Fig. 2A. —Transverse T1-weighted three-dimensional spoiled gradient-echo MR images of 53-year-old man with cirrhosis and hepatitis B. Biopsy of suspicious liver mass revealed hepatocellular carcinoma. Contrast-enhanced arterial phase image obtained through left portal vein (arrowhead) reveals arterially enhancing mass (arrow) in anterior segment of right lobe of liver.

 


View larger version (130K):

[in a new window]
  		Fig. 2B. —Transverse T1-weighted three-dimensional spoiled gradient-echo MR images of 53-year-old man with cirrhosis and hepatitis B. Biopsy of suspicious liver mass revealed hepatocellular carcinoma. Portal venous phase image obtained at same level as A shows venous washout of contrast material within mass (arrow). Arrowhead = left portal vein.

 


View larger version (117K):

[in a new window]
  		Fig. 3. —Graph shows estimated probability of hepatic malignancy using presence or absence of arterial phase enhancement and of venous phase washout and serum {alpha}-fetoprotein level. With baseline patient age of 55 years, index lesion size of 4.4 cm, and three lesions, probability of malignancy exceeds 60% regardless of {alpha}-fetoprotein level if arterial phase enhancement and either venous or delayed phase washout are present ({square}); it rises to more than 80% if {alpha}-fetoprotein value is greater than 10 ng/mL. In patients with a normal {alpha}-fetoprotein level (<20 ng/mL), probability of malignancy is less than 10% if both arterial phase enhancement and venous phase washout are absent ({diamond}). In average patient with either arterial enhancement and no venous washout ({Delta}) or no arterial enhancement and venous washout ({circ}), but not both, contribution of {alpha}-fetoprotein level to estimation of malignancy is increased.

 


View larger version (17K):

[in a new window]
  		Fig. 4A. —Transverse T1-weighted three-dimensional spoiled gradient-echo MR images of 64-year-old man with cirrhosis and hepatitis C. Biopsy of suspicious liver lesion revealed cirrhosis without evidence of malignancy. Contrast-enhanced arterial phase image shows faintly enhancing mass (arrow) in dome of liver.

 


View larger version (127K):

[in a new window]
  		Fig. 4B. —Transverse T1-weighted three-dimensional spoiled gradient-echo MR images of 64-year-old man with cirrhosis and hepatitis C. Biopsy of suspicious liver lesion revealed cirrhosis without evidence of malignancy. Portal venous phase image obtained at same level as A reveals persistent enhancement of mass (arrow) compared with rest of liver.

 


View larger version (13K):

[in a new window]
  		Fig. 5. —Graph shows independent effects of number of lesions (one lesion, •; three lesions, {blacksquare}; five lesions, {blacktriangleup}) on probability of hepatic malignancy. In presence of arterial phase enhancement and venous phase washout, number of lesions present has marked effect on probability of malignancy; however, as {alpha}-fetoprotein level rises above 20 ng/mL, probability of malignancy exceeds 70% regardless how many lesions are present.

 

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Copyright © 2003 by the American Roentgen Ray Society.