Non-redundant Combination of Hand-Crafted and Deep Learning Radiomics: Application to the Early Detection of Pancreatic Cancer

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

doi:10.1007/978-3-031-45350-2_6

Non-redundant Combination of Hand-Crafted and Deep Learning Radiomics: Application to the Early Detection of Pancreatic Cancer

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

Institut Polytechnique de Paris
Guerbet Research
Hospital of Annecy-Genevois
Laboratoire de Probabilités et Modèles Aléatoires
Sorbonne Université

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

3 Citations (Scopus)

Abstract

We address the problem of learning Deep Learning Radiomics (DLR) that are not redundant with Hand-Crafted Radiomics (HCR). To do so, we extract DLR features using a VAE while enforcing their independence with HCR features by minimizing their mutual information. The resulting DLR features can be combined with hand-crafted ones and leveraged by a classifier to predict early markers of cancer. We illustrate our method on four early markers of pancreatic cancer and validate it on a large independent test set. Our results highlight the value of combining non-redundant DLR and HCR features, as evidenced by an improvement in the Area Under the Curve compared to baseline methods that do not address redundancy or solely rely on HCR features.

Original language	English
Title of host publication	Cancer Prevention Through Early Detection - 2nd International Workshop, CaPTion 2023, Held in Conjunction with MICCAI 2023, Proceedings
Editors	Sharib Ali, Fons van der Sommen, Maureen van Eijnatten, Iris Kolenbrander, Bartłomiej W. Papież, Yueming Jin
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	68-82
Number of pages	15
ISBN (Print)	9783031453496
DOIs	https://doi.org/10.1007/978-3-031-45350-2_6
Publication status	Published - 1 Jan 2023
Event	2nd International Workshop on Cancer Prevention through early detecTion, CaPTion 2023 - Vancover, Canada Duration: 12 Oct 2023 → 12 Oct 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	14295 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	2nd International Workshop on Cancer Prevention through early detecTion, CaPTion 2023
Country/Territory	Canada
City	Vancover
Period	12/10/23 → 12/10/23

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Early Diagnosis
Mutual Information
Pancreatic Cancer
Radiomics
Variational Autoencoders

Access to Document

10.1007/978-3-031-45350-2_6

Cite this

Vétil, R., Abi-Nader, C., Bône, A., Vullierme, M. P., Rohé, M. M., Gori, P., & Bloch, I. (2023). Non-redundant Combination of Hand-Crafted and Deep Learning Radiomics: Application to the Early Detection of Pancreatic Cancer. In S. Ali, F. van der Sommen, M. van Eijnatten, I. Kolenbrander, B. W. Papież, & Y. Jin (Eds.), Cancer Prevention Through Early Detection - 2nd International Workshop, CaPTion 2023, Held in Conjunction with MICCAI 2023, Proceedings (pp. 68-82). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14295 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-45350-2_6

Vétil, Rebeca ; Abi-Nader, Clément ; Bône, Alexandre et al. / Non-redundant Combination of Hand-Crafted and Deep Learning Radiomics : Application to the Early Detection of Pancreatic Cancer. Cancer Prevention Through Early Detection - 2nd International Workshop, CaPTion 2023, Held in Conjunction with MICCAI 2023, Proceedings. editor / Sharib Ali ; Fons van der Sommen ; Maureen van Eijnatten ; Iris Kolenbrander ; Bartłomiej W. Papież ; Yueming Jin. Springer Science and Business Media Deutschland GmbH, 2023. pp. 68-82 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{013701dc46314a27bc552377c8e49e66,

title = "Non-redundant Combination of Hand-Crafted and Deep Learning Radiomics: Application to the Early Detection of Pancreatic Cancer",

abstract = "We address the problem of learning Deep Learning Radiomics (DLR) that are not redundant with Hand-Crafted Radiomics (HCR). To do so, we extract DLR features using a VAE while enforcing their independence with HCR features by minimizing their mutual information. The resulting DLR features can be combined with hand-crafted ones and leveraged by a classifier to predict early markers of cancer. We illustrate our method on four early markers of pancreatic cancer and validate it on a large independent test set. Our results highlight the value of combining non-redundant DLR and HCR features, as evidenced by an improvement in the Area Under the Curve compared to baseline methods that do not address redundancy or solely rely on HCR features.",

keywords = "Early Diagnosis, Mutual Information, Pancreatic Cancer, Radiomics, Variational Autoencoders",

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} 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.; 2nd International Workshop on Cancer Prevention through early detecTion, CaPTion 2023 ; Conference date: 12-10-2023 Through 12-10-2023",

year = "2023",

month = jan,

day = "1",

doi = "10.1007/978-3-031-45350-2\_6",

language = "English",

isbn = "9783031453496",

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 = "68--82",

editor = "Sharib Ali and \{van der Sommen\}, Fons and \{van Eijnatten\}, Maureen and Iris Kolenbrander and Papie{\.z}, \{Bart{\l}omiej W.\} and Yueming Jin",

booktitle = "Cancer Prevention Through Early Detection - 2nd International Workshop, CaPTion 2023, Held in Conjunction with MICCAI 2023, Proceedings",

}

Vétil, R, Abi-Nader, C, Bône, A, Vullierme, MP, Rohé, MM, Gori, P & Bloch, I 2023, Non-redundant Combination of Hand-Crafted and Deep Learning Radiomics: Application to the Early Detection of Pancreatic Cancer. in S Ali, F van der Sommen, M van Eijnatten, I Kolenbrander, BW Papież & Y Jin (eds), Cancer Prevention Through Early Detection - 2nd International Workshop, CaPTion 2023, Held in Conjunction with MICCAI 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14295 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 68-82, 2nd International Workshop on Cancer Prevention through early detecTion, CaPTion 2023, Vancover, Canada, 12/10/23. https://doi.org/10.1007/978-3-031-45350-2_6

Non-redundant Combination of Hand-Crafted and Deep Learning Radiomics: Application to the Early Detection of Pancreatic Cancer. / Vétil, Rebeca; Abi-Nader, Clément; Bône, Alexandre et al.
Cancer Prevention Through Early Detection - 2nd International Workshop, CaPTion 2023, Held in Conjunction with MICCAI 2023, Proceedings. ed. / Sharib Ali; Fons van der Sommen; Maureen van Eijnatten; Iris Kolenbrander; Bartłomiej W. Papież; Yueming Jin. Springer Science and Business Media Deutschland GmbH, 2023. p. 68-82 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14295 LNCS).

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

TY - GEN

T1 - Non-redundant Combination of Hand-Crafted and Deep Learning Radiomics

T2 - 2nd International Workshop on Cancer Prevention through early detecTion, CaPTion 2023

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 - 2023/1/1

Y1 - 2023/1/1

N2 - We address the problem of learning Deep Learning Radiomics (DLR) that are not redundant with Hand-Crafted Radiomics (HCR). To do so, we extract DLR features using a VAE while enforcing their independence with HCR features by minimizing their mutual information. The resulting DLR features can be combined with hand-crafted ones and leveraged by a classifier to predict early markers of cancer. We illustrate our method on four early markers of pancreatic cancer and validate it on a large independent test set. Our results highlight the value of combining non-redundant DLR and HCR features, as evidenced by an improvement in the Area Under the Curve compared to baseline methods that do not address redundancy or solely rely on HCR features.

AB - We address the problem of learning Deep Learning Radiomics (DLR) that are not redundant with Hand-Crafted Radiomics (HCR). To do so, we extract DLR features using a VAE while enforcing their independence with HCR features by minimizing their mutual information. The resulting DLR features can be combined with hand-crafted ones and leveraged by a classifier to predict early markers of cancer. We illustrate our method on four early markers of pancreatic cancer and validate it on a large independent test set. Our results highlight the value of combining non-redundant DLR and HCR features, as evidenced by an improvement in the Area Under the Curve compared to baseline methods that do not address redundancy or solely rely on HCR features.

KW - Early Diagnosis

KW - Mutual Information

KW - Pancreatic Cancer

KW - Radiomics

KW - Variational Autoencoders

U2 - 10.1007/978-3-031-45350-2_6

DO - 10.1007/978-3-031-45350-2_6

M3 - Conference contribution

AN - SCOPUS:85175983539

SN - 9783031453496

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

SP - 68

EP - 82

BT - Cancer Prevention Through Early Detection - 2nd International Workshop, CaPTion 2023, Held in Conjunction with MICCAI 2023, Proceedings

A2 - Ali, Sharib

A2 - van der Sommen, Fons

A2 - van Eijnatten, Maureen

A2 - Kolenbrander, Iris

A2 - Papież, Bartłomiej W.

A2 - Jin, Yueming

PB - Springer Science and Business Media Deutschland GmbH

Y2 - 12 October 2023 through 12 October 2023

ER -

Vétil R, Abi-Nader C, Bône A, Vullierme MP, Rohé MM, Gori P et al. Non-redundant Combination of Hand-Crafted and Deep Learning Radiomics: Application to the Early Detection of Pancreatic Cancer. In Ali S, van der Sommen F, van Eijnatten M, Kolenbrander I, Papież BW, Jin Y, editors, Cancer Prevention Through Early Detection - 2nd International Workshop, CaPTion 2023, Held in Conjunction with MICCAI 2023, Proceedings. Springer Science and Business Media Deutschland GmbH. 2023. p. 68-82. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-45350-2_6

Non-redundant Combination of Hand-Crafted and Deep Learning Radiomics: Application to the Early Detection of Pancreatic Cancer

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Fingerprint

Cite this