Supervised in-vivo plaque characterization incorporating class label uncertainty

Arna Van Engelen*, Wiro J. Niessen, Stefan Klein, Harald C. Groen, Hence J.M. Verhagen, Jolanda J. Wentzel, Aad Van Der Lugt, Marleen De Bruijne

*Corresponding author for this work

Research output: Chapter/Conference proceedingConference proceedingAcademicpeer-review

6 Citations (Scopus)

Abstract

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%.

Original languageEnglish
Title of host publication2012 9th IEEE International Symposium on Biomedical Imaging
Subtitle of host publicationFrom Nano to Macro, ISBI 2012 - Proceedings
Pages246-249
Number of pages4
DOIs
Publication statusPublished - 12 Jul 2012
Event2012 9th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2012 - Barcelona, Spain
Duration: 2 May 20125 May 2012

Publication series

SeriesProceedings - International Symposium on Biomedical Imaging
ISSN1945-7928

Conference

Conference2012 9th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2012
Country/TerritorySpain
CityBarcelona
Period2/05/125/05/12

Fingerprint

Dive into the research topics of 'Supervised in-vivo plaque characterization incorporating class label uncertainty'. Together they form a unique fingerprint.

Cite this