Supervised in-vivo plaque characterization incorporating class label uncertainty

TY - GEN

T1 - Supervised in-vivo plaque characterization incorporating class label uncertainty

AU - Engelen, Arna van

AU - Niessen, Wiro J.

AU - Klein, Stefan

AU - Groen, Harald C.

AU - Verhagen, Hence J. M.

AU - Wentzel, Jolanda J.

AU - van der Lugt, Aad

AU - de Bruijne, Marleen

N1 - Conference code: 9

PY - 2012

Y1 - 2012

N2 - 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%.

AB - 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%.

U2 - 10.1109/ISBI.2012.6235530

DO - 10.1109/ISBI.2012.6235530

M3 - Article in proceedings

SN - 978-1-4577-1858-8

SP - 246

EP - 249

BT - 2012 9th IEEE International Symposium on Biomedical Imaging (ISBI)

PB - IEEE

T2 - 2012 9th IEEE International Symposium on Biomedical Imaging: From Nano to Macro

Y2 - 2 May 2012 through 5 May 2012

ER -