DocumentCode :
1123492
Title :
Hepatic Perfusion Imaging Using Factor Analysis of Contrast Enhanced Ultrasound
Author :
Lueck, G.J. ; Kim, T.K. ; Burns, P.N. ; Martel, A.L.
Author_Institution :
Res. Inst. & Dept. of Med. Biophys., Univ. of Toronto, Toronto, ON
Volume :
27
Issue :
10
fYear :
2008
Firstpage :
1449
Lastpage :
1457
Abstract :
Contrast enhanced ultrasound imaging provides a real-time tool for evaluating vasculature in the liver. Primary liver cancer is known to be perfused exclusively by blood from the hepatic artery, whereas normal liver is also supplied by the portal vein. Visual separation of two different phases of enhancement from the independent feeding vessels is important for diagnosis but remains a challenge. This paper presents a method of using factor analysis for extracting distinct time-intensity curves. A key component to this extraction is the clustering of measured bolus curves and their projection onto a positivity domain to obtain nonnegative curves. This technique provides complementary images representing spatial loadings on each curve. As little as 1% of the data is required to contain unmixed signals to extract time-intensity curves that correlate well with true curves. A method of combining this information to display a regional hepatic perfusion image is proposed, and results are tested on a set of 10 patients. Region of interest analysis suggests it is possible to detect changes in the hepatic perfusion index of liver lesions relative to normal liver parenchyma using contrast ultrasound.
Keywords :
biomedical ultrasonics; blood vessels; cancer; haemorheology; image enhancement; liver; medical image processing; principal component analysis; tumours; complementary images; contrast enhanced ultrasound imaging; distinct time-intensity curves; factor analysis; hepatic arteries; hepatic perfusion index; liver lesions; liver vasculature evaluation; measured bolus curves; normal liver parenchyma; portal vein; primary liver cancer; principal components analysis; regional hepatic perfusion imaging; spatial loadings; visual separation; Arteries; Blood; Cancer; Data mining; Image analysis; Independent component analysis; Liver; Portals; Ultrasonic imaging; Veins; Factor analysis; Ultrasound; factor analysis; hepatic perfusion index; microbubbles; ultrasound; Algorithms; Artificial Intelligence; Contrast Media; Factor Analysis, Statistical; Fluorocarbons; Hepatic Artery; Humans; Image Enhancement; Image Interpretation, Computer-Assisted; Microbubbles; Multivariate Analysis; Pattern Recognition, Automated; Phantoms, Imaging; Reproducibility of Results; Sensitivity and Specificity; Ultrasonography;
fLanguage :
English
Journal_Title :
Medical Imaging, IEEE Transactions on
Publisher :
ieee
ISSN :
0278-0062
Type :
jour
DOI :
10.1109/TMI.2008.922695
Filename :
4483772
Link To Document :
بازگشت