Uncertainty Quantification in CNN-Based Surface Prediction Using Shape Priors

TY - JOUR

T1 - Uncertainty Quantification in CNN-Based Surface Prediction Using Shape Priors

AU - Tóthová, Katarína

AU - Parisot, Sarah

AU - Lee, Matthew C.H.

AU - Puyol-Antón, Esther

AU - Koch, Lisa M.

AU - King, Andrew P.

AU - Konukoglu, Ender

AU - Pollefeys, Marc

PY - 2018/11/23

Y1 - 2018/11/23

N2 - Surface reconstruction is a vital tool in a wide range of areas of medical image analysis and clinical research. Despite the fact that many methods have proposed solutions to the reconstruction problem, most, due to their deterministic nature, do not directly address the issue of quantifying uncertainty associated with their predictions. We remedy this by proposing a novel probabilistic deep learning approach capable of simultaneous surface reconstruction and associated uncertainty prediction. The method incorporates prior shape information in the form of a principal component analysis (PCA) model. Experiments using the UK Biobank data show that our probabilistic approach outperforms an analogous deterministic PCA-based method in the task of 2D organ delineation and quantifies uncertainty by formulating distributions over predicted surface vertex positions.

AB - Surface reconstruction is a vital tool in a wide range of areas of medical image analysis and clinical research. Despite the fact that many methods have proposed solutions to the reconstruction problem, most, due to their deterministic nature, do not directly address the issue of quantifying uncertainty associated with their predictions. We remedy this by proposing a novel probabilistic deep learning approach capable of simultaneous surface reconstruction and associated uncertainty prediction. The method incorporates prior shape information in the form of a principal component analysis (PCA) model. Experiments using the UK Biobank data show that our probabilistic approach outperforms an analogous deterministic PCA-based method in the task of 2D organ delineation and quantifies uncertainty by formulating distributions over predicted surface vertex positions.

KW - Deep learning

KW - Shape prior

KW - Surface reconstruction

KW - Uncertainty quantification

UR - http://www.scopus.com/inward/record.url?scp=85057381858&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-04747-4_28

DO - 10.1007/978-3-030-04747-4_28

M3 - Conference paper

AN - SCOPUS:85057381858

SN - 0302-9743

SP - 300

EP - 310

JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

T2 - International Workshop on Shape in Medical Imaging, ShapeMI 2018 held in conjunction with 21st International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2018

Y2 - 20 September 2018 through 20 September 2018

ER -