Learning Shape Distributions from Large Databases of Healthy Organs: Applications to Zero-Shot and Few-Shot Abnormal Pancreas Detection

Rebeca Vétil; Clément Abi-Nader; Alexandre Bône; Marie Pierre Vullierme; Marc Michel Rohé; Pietro Gori; Isabelle Bloch

doi:10.1007/978-3-031-16434-7_45

Learning Shape Distributions from Large Databases of Healthy Organs: Applications to Zero-Shot and Few-Shot Abnormal Pancreas Detection

Rebeca Vétil
, Clément Abi-Nader
, Alexandre Bône
, Marie Pierre Vullierme
, Marc Michel Rohé
, Pietro Gori
, Isabelle Bloch

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

We propose a scalable and data-driven approach to learn shape distributions from large databases of healthy organs. To do so, volumetric segmentation masks are embedded into a common probabilistic shape space that is learned with a variational auto-encoding network. The resulting latent shape representations are leveraged to derive zero-shot and few-shot methods for abnormal shape detection. The proposed distribution learning approach is illustrated on a large database of 1200 healthy pancreas shapes. Downstream qualitative and quantitative experiments are conducted on a separate test set of 224 pancreas from patients with mixed conditions. The abnormal pancreas detection AUC reached up to 65.41 % in the zero-shot configuration, and 78.97 % in the few-shot configuration with as few as 15 abnormal examples, outperforming a baseline approach based on the sole volume.

Original language	English
Title of host publication	Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference, Proceedings
Editors	Linwei Wang, Qi Dou, P. Thomas Fletcher, Stefanie Speidel, Shuo Li
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	464-473
Number of pages	10
ISBN (Print)	9783031164330
DOIs	https://doi.org/10.1007/978-3-031-16434-7_45
Publication status	Published - 1 Jan 2022
Event	25th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2022 - Singapore, Singapore Duration: 18 Sept 2022 → 22 Sept 2022

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	13432 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	25th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2022
Country/Territory	Singapore
City	Singapore
Period	18/09/22 → 22/09/22

Keywords

Anomaly detection
Pancreas
Shape analysis

Access to Document

10.1007/978-3-031-16434-7_45

Cite this

Vétil, R., Abi-Nader, C., Bône, A., Vullierme, M. P., Rohé, M. M., Gori, P., & Bloch, I. (2022). Learning Shape Distributions from Large Databases of Healthy Organs: Applications to Zero-Shot and Few-Shot Abnormal Pancreas Detection. In L. Wang, Q. Dou, P. T. Fletcher, S. Speidel, & S. Li (Eds.), Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference, Proceedings (pp. 464-473). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13432 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-16434-7_45

Vétil, Rebeca ; Abi-Nader, Clément ; Bône, Alexandre et al. / Learning Shape Distributions from Large Databases of Healthy Organs : Applications to Zero-Shot and Few-Shot Abnormal Pancreas Detection. Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference, Proceedings. editor / Linwei Wang ; Qi Dou ; P. Thomas Fletcher ; Stefanie Speidel ; Shuo Li. Springer Science and Business Media Deutschland GmbH, 2022. pp. 464-473 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{201a7cef091b489b9e847cb0c385464d,

title = "Learning Shape Distributions from Large Databases of Healthy Organs: Applications to Zero-Shot and Few-Shot Abnormal Pancreas Detection",

abstract = "We propose a scalable and data-driven approach to learn shape distributions from large databases of healthy organs. To do so, volumetric segmentation masks are embedded into a common probabilistic shape space that is learned with a variational auto-encoding network. The resulting latent shape representations are leveraged to derive zero-shot and few-shot methods for abnormal shape detection. The proposed distribution learning approach is illustrated on a large database of 1200 healthy pancreas shapes. Downstream qualitative and quantitative experiments are conducted on a separate test set of 224 pancreas from patients with mixed conditions. The abnormal pancreas detection AUC reached up to 65.41 \% in the zero-shot configuration, and 78.97 \% in the few-shot configuration with as few as 15 abnormal examples, outperforming a baseline approach based on the sole volume.",

keywords = "Anomaly detection, Pancreas, Shape analysis",

author = "Rebeca V{\'e}til and Cl{\'e}ment Abi-Nader and Alexandre B{\^o}ne and Vullierme, \{Marie Pierre\} and Roh{\'e}, \{Marc Michel\} and Pietro Gori and Isabelle Bloch",

note = "Publisher Copyright: {\textcopyright} 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.; 25th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2022 ; Conference date: 18-09-2022 Through 22-09-2022",

year = "2022",

month = jan,

day = "1",

doi = "10.1007/978-3-031-16434-7\_45",

language = "English",

isbn = "9783031164330",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "464--473",

editor = "Linwei Wang and Qi Dou and Fletcher, \{P. Thomas\} and Stefanie Speidel and Shuo Li",

booktitle = "Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference, Proceedings",

}

Vétil, R, Abi-Nader, C, Bône, A, Vullierme, MP, Rohé, MM, Gori, P & Bloch, I 2022, Learning Shape Distributions from Large Databases of Healthy Organs: Applications to Zero-Shot and Few-Shot Abnormal Pancreas Detection. in L Wang, Q Dou, PT Fletcher, S Speidel & S Li (eds), Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 13432 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 464-473, 25th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2022, Singapore, Singapore, 18/09/22. https://doi.org/10.1007/978-3-031-16434-7_45

Learning Shape Distributions from Large Databases of Healthy Organs: Applications to Zero-Shot and Few-Shot Abnormal Pancreas Detection. / Vétil, Rebeca; Abi-Nader, Clément; Bône, Alexandre et al.
Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference, Proceedings. ed. / Linwei Wang; Qi Dou; P. Thomas Fletcher; Stefanie Speidel; Shuo Li. Springer Science and Business Media Deutschland GmbH, 2022. p. 464-473 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13432 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Learning Shape Distributions from Large Databases of Healthy Organs

T2 - 25th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2022

AU - Vétil, Rebeca

AU - Abi-Nader, Clément

AU - Bône, Alexandre

AU - Vullierme, Marie Pierre

AU - Rohé, Marc Michel

AU - Gori, Pietro

AU - Bloch, Isabelle

PY - 2022/1/1

Y1 - 2022/1/1

N2 - We propose a scalable and data-driven approach to learn shape distributions from large databases of healthy organs. To do so, volumetric segmentation masks are embedded into a common probabilistic shape space that is learned with a variational auto-encoding network. The resulting latent shape representations are leveraged to derive zero-shot and few-shot methods for abnormal shape detection. The proposed distribution learning approach is illustrated on a large database of 1200 healthy pancreas shapes. Downstream qualitative and quantitative experiments are conducted on a separate test set of 224 pancreas from patients with mixed conditions. The abnormal pancreas detection AUC reached up to 65.41 % in the zero-shot configuration, and 78.97 % in the few-shot configuration with as few as 15 abnormal examples, outperforming a baseline approach based on the sole volume.

AB - We propose a scalable and data-driven approach to learn shape distributions from large databases of healthy organs. To do so, volumetric segmentation masks are embedded into a common probabilistic shape space that is learned with a variational auto-encoding network. The resulting latent shape representations are leveraged to derive zero-shot and few-shot methods for abnormal shape detection. The proposed distribution learning approach is illustrated on a large database of 1200 healthy pancreas shapes. Downstream qualitative and quantitative experiments are conducted on a separate test set of 224 pancreas from patients with mixed conditions. The abnormal pancreas detection AUC reached up to 65.41 % in the zero-shot configuration, and 78.97 % in the few-shot configuration with as few as 15 abnormal examples, outperforming a baseline approach based on the sole volume.

KW - Anomaly detection

KW - Pancreas

KW - Shape analysis

U2 - 10.1007/978-3-031-16434-7_45

DO - 10.1007/978-3-031-16434-7_45

M3 - Conference contribution

AN - SCOPUS:85139033776

SN - 9783031164330

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 464

EP - 473

BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference, Proceedings

A2 - Wang, Linwei

A2 - Dou, Qi

A2 - Fletcher, P. Thomas

A2 - Speidel, Stefanie

A2 - Li, Shuo

PB - Springer Science and Business Media Deutschland GmbH

Y2 - 18 September 2022 through 22 September 2022

ER -

Vétil R, Abi-Nader C, Bône A, Vullierme MP, Rohé MM, Gori P et al. Learning Shape Distributions from Large Databases of Healthy Organs: Applications to Zero-Shot and Few-Shot Abnormal Pancreas Detection. In Wang L, Dou Q, Fletcher PT, Speidel S, Li S, editors, Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference, Proceedings. Springer Science and Business Media Deutschland GmbH. 2022. p. 464-473. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-16434-7_45