TY - GEN
T1 - Probabilistic Point Cloud Reconstructions for Vertebral Shape Analysis
AU - Sekuboyina, Anjany
AU - Rempfler, Markus
AU - Valentinitsch, Alexander
AU - Loeffler, Maximilian
AU - Kirschke, Jan S.
AU - Menze, Bjoern H.
N1 - Publisher Copyright:
© 2019, Springer Nature Switzerland AG.
PY - 2019
Y1 - 2019
N2 - We propose an auto-encoding network architecture for point clouds (PC) capable of extracting shape signatures without supervision. Building on this, we (i) design a loss function capable of modelling data variance on PCs which are unstructured, and (ii) regularise the latent space as in a variational auto-encoder, both of which increase the auto-encoders’ descriptive capacity while making them probabilistic. Evaluating the reconstruction quality of our architectures, we employ them for detecting vertebral fractures without any supervision. By learning to efficiently reconstruct only healthy vertebrae, fractures are detected as anomalous reconstructions. Evaluating on a dataset containing ∼ 1500 vertebrae, we achieve area-under-ROC curve of >75%, without using intensity-based features.
AB - We propose an auto-encoding network architecture for point clouds (PC) capable of extracting shape signatures without supervision. Building on this, we (i) design a loss function capable of modelling data variance on PCs which are unstructured, and (ii) regularise the latent space as in a variational auto-encoder, both of which increase the auto-encoders’ descriptive capacity while making them probabilistic. Evaluating the reconstruction quality of our architectures, we employ them for detecting vertebral fractures without any supervision. By learning to efficiently reconstruct only healthy vertebrae, fractures are detected as anomalous reconstructions. Evaluating on a dataset containing ∼ 1500 vertebrae, we achieve area-under-ROC curve of >75%, without using intensity-based features.
UR - http://www.scopus.com/inward/record.url?scp=85075818945&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-32226-7_42
DO - 10.1007/978-3-030-32226-7_42
M3 - Conference contribution
AN - SCOPUS:85075818945
SN - 9783030322250
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 375
EP - 383
BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2019 - 22nd International Conference, Proceedings
A2 - Shen, Dinggang
A2 - Yap, Pew-Thian
A2 - Liu, Tianming
A2 - Peters, Terry M.
A2 - Khan, Ali
A2 - Staib, Lawrence H.
A2 - Essert, Caroline
A2 - Zhou, Sean
PB - Springer Science and Business Media Deutschland GmbH
T2 - 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019
Y2 - 13 October 2019 through 17 October 2019
ER -