Supervised in-vivo plaque characterization incorporating class label uncertainty

• Published in: 2012 9th IEEE International Symposium on Biomedical Imaging (ISBI) pp. 246 - 249
• Main Authors: van Engelen, Arna, Niessen, Wiro J, Klein, Stefan, Groen, Harald C, Verhagen, Hence J M, Wentzel, Jolanda J, van der Lugt, Aad, de Bruijne, Marleen
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.05.2012
• Subjects: Accuracy; Atherosclerosis; classification; Computed tomography; Correlation; histology; Image segmentation; Magnetic resonance imaging; pattern recognition; Segmentation; Training
• ISBN: 145771857X; 9781457718571
• ISSN: 1945-7928
• DOI: 10.1109/ISBI.2012.6235530

We segment atherosclerotic plaque components in in-vivo MRI and CT data using supervised voxelwise classification. The most reliable ground truth can be obtained from histology sections, however, it is not straightforward to use this for classifier training as the registration with in-vivo data often shows misalignments. Therefore, for training we incorporate uncertainty in the ground truth via "soft" labels that indicate a probability for each class. Soft labels are created by Gaussian blurring of the original "hard" segmentations, and weighted by the registration accuracy. Classification is evaluated on the relative volumes for fibrous, lipid-rich necrotic and calcified tissue. Using conventional hard labels, the differences between the ground truth and classification result per subject are −0.4±3.6% for calcification, +7.6±14.9% for fibrous and −7.2±14.5% for necrotic tissue. Using the new approach accuracy is improved: for calcification −0.6±1.6%, fibrous +3.6±16.8% and necrotic tissue −2.9±16.1%.