DocumentCode
3173745
Title
Contour detection of short axis slice MR images for contraction irregularity assessment
Author
Muller, A. ; Neitmann, A. ; Merkle, N. ; Wohrle, J. ; Hombach, V. ; Kestler, Ha
Author_Institution
Internal Medicine I, Univ. Hosp. Ulm
fYear
2005
fDate
25-28 Sept. 2005
Firstpage
21
Lastpage
24
Abstract
A new method for automatically detecting left ventricular endocardial borders in cardiac MRI short-axis slices was applied to a previously published patient group for which manually drawn contours already existed. An initial point lying in the left ventricular endocardium was found by a size invariant circular Hough transform. A coarse segmentation of the endocardial region was completed by merging surrounding regions and building the convex hull. Adjacent time frames were used to remove jitter from the contour movements. A feature (IRM), extracted from those contours, quantified the irregularities of left ventricular contractions. Patients with dilated cardiomyopathy were separated from the control group by thresholding the IRM parameter
Keywords
Hough transforms; biomedical MRI; cardiology; diseases; edge detection; feature extraction; image segmentation; medical image processing; cardiac MRI short-axis slices; coarse endocardial segmentation; contour detection; contour movement; contraction irregularity assessment; dilated cardiomyopathy; feature extraction; irregularities quantification; jitter removal; left ventricular contraction; left ventricular endocardial border detection; size invariant circular Hough transform; Biomedical imaging; Cardiology; Feature extraction; Heart; Hospitals; Jitter; Magnetic resonance imaging; Merging; Pacemakers; Phase measurement;
fLanguage
English
Publisher
ieee
Conference_Titel
Computers in Cardiology, 2005
Conference_Location
Lyon
Print_ISBN
0-7803-9337-6
Type
conf
DOI
10.1109/CIC.2005.1588023
Filename
1588023
Link To Document