A radiomics approach for automated identification of aggressive tumors on combined PET and multi-parametric MRI

Tao Wan; Bixiao Cui; Yaping Wang; Zengchang Qin; Jie Lu

doi:10.1007/978-3-319-70136-3_77

A radiomics approach for automated identification of aggressive tumors on combined PET and multi-parametric MRI

Tao Wan^*
, Bixiao Cui
, Yaping Wang
, Zengchang Qin
, Jie Lu

^*Corresponding author for this work

Capital Medical University
Beihang University

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

1 Scopus citations

Abstract

We present a computerized image-based method to automatically identify aggressive tumors on combined positron emission tomography and magnetic resonance imaging (PET-MRI) using radiomics texture features from both PET and multi-parametric MRI (MP-MRI). The work aims at investigating the potential use of new composite textures from PET-MRI for the assessment of different biological properties present in cancer and non-cancer regions, and eventually for early detection of malignant tumors in real clinical practice. Towards this goal, a large number of radiomics features are extracted to characterize the intratumoural heterogeneity and microarchitectural morphologic differences within tumors. These image attributes are valuable for determining tumor aggressiveness. The radiomics model was evaluated on three types of cancers (pancreas, gallbladder, and liver). Compared to single image modality (PET or MRI), the fused PET and MP-MRI achieved the best classification performance in differentiating cancer and non-cancer regions with the area of under curve (AUC) of 0.87 for pancreas cancer, 0.89 for gallbladder cancer, and 0.82 for liver cancer. The results indicated that PET-MRI based imaging biomarkers could be useful in identifying aggressive tumors.

Original language	English
Title of host publication	Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings
Editors	Derong Liu, Shengli Xie, Dongbin Zhao, Yuanqing Li, El-Sayed M. El-Alfy
Publisher	Springer Verlag
Pages	731-739
Number of pages	9
ISBN (Print)	9783319701356
DOIs	https://doi.org/10.1007/978-3-319-70136-3_77
State	Published - 2017
Event	24th International Conference on Neural Information Processing, ICONIP 2017 - Guangzhou, China Duration: 14 Nov 2017 → 18 Nov 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10639 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	24th International Conference on Neural Information Processing, ICONIP 2017
Country/Territory	China
City	Guangzhou
Period	14/11/17 → 18/11/17

UN SDGs

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

SDG 3 Good Health and Well-being

Keywords

Aggressive tumor
Multi-parametric MRI
PET
Radiomics

Access to Document

10.1007/978-3-319-70136-3_77

Cite this

Wan, T., Cui, B., Wang, Y., Qin, Z., & Lu, J. (2017). A radiomics approach for automated identification of aggressive tumors on combined PET and multi-parametric MRI. In D. Liu, S. Xie, D. Zhao, Y. Li, & E.-S. M. El-Alfy (Eds.), Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings (pp. 731-739). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10639 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-70136-3_77

Wan, Tao ; Cui, Bixiao ; Wang, Yaping et al. / A radiomics approach for automated identification of aggressive tumors on combined PET and multi-parametric MRI. Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings. editor / Derong Liu ; Shengli Xie ; Dongbin Zhao ; Yuanqing Li ; El-Sayed M. El-Alfy. Springer Verlag, 2017. pp. 731-739 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{8a2d52828a85468c9dc108f36dd21b4f,

title = "A radiomics approach for automated identification of aggressive tumors on combined PET and multi-parametric MRI",

abstract = "We present a computerized image-based method to automatically identify aggressive tumors on combined positron emission tomography and magnetic resonance imaging (PET-MRI) using radiomics texture features from both PET and multi-parametric MRI (MP-MRI). The work aims at investigating the potential use of new composite textures from PET-MRI for the assessment of different biological properties present in cancer and non-cancer regions, and eventually for early detection of malignant tumors in real clinical practice. Towards this goal, a large number of radiomics features are extracted to characterize the intratumoural heterogeneity and microarchitectural morphologic differences within tumors. These image attributes are valuable for determining tumor aggressiveness. The radiomics model was evaluated on three types of cancers (pancreas, gallbladder, and liver). Compared to single image modality (PET or MRI), the fused PET and MP-MRI achieved the best classification performance in differentiating cancer and non-cancer regions with the area of under curve (AUC) of 0.87 for pancreas cancer, 0.89 for gallbladder cancer, and 0.82 for liver cancer. The results indicated that PET-MRI based imaging biomarkers could be useful in identifying aggressive tumors.",

keywords = "Aggressive tumor, Multi-parametric MRI, PET, Radiomics",

author = "Tao Wan and Bixiao Cui and Yaping Wang and Zengchang Qin and Jie Lu",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2017.; 24th International Conference on Neural Information Processing, ICONIP 2017 ; Conference date: 14-11-2017 Through 18-11-2017",

year = "2017",

doi = "10.1007/978-3-319-70136-3\_77",

language = "英语",

isbn = "9783319701356",

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

publisher = "Springer Verlag",

pages = "731--739",

editor = "Derong Liu and Shengli Xie and Dongbin Zhao and Yuanqing Li and El-Alfy, \{El-Sayed M.\}",

booktitle = "Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings",

address = "德国",

}

Wan, T, Cui, B, Wang, Y, Qin, Z & Lu, J 2017, A radiomics approach for automated identification of aggressive tumors on combined PET and multi-parametric MRI. in D Liu, S Xie, D Zhao, Y Li & E-SM El-Alfy (eds), Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10639 LNCS, Springer Verlag, pp. 731-739, 24th International Conference on Neural Information Processing, ICONIP 2017, Guangzhou, China, 14/11/17. https://doi.org/10.1007/978-3-319-70136-3_77

A radiomics approach for automated identification of aggressive tumors on combined PET and multi-parametric MRI. / Wan, Tao; Cui, Bixiao; Wang, Yaping et al.
Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings. ed. / Derong Liu; Shengli Xie; Dongbin Zhao; Yuanqing Li; El-Sayed M. El-Alfy. Springer Verlag, 2017. p. 731-739 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10639 LNCS).

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

TY - GEN

T1 - A radiomics approach for automated identification of aggressive tumors on combined PET and multi-parametric MRI

AU - Wan, Tao

AU - Cui, Bixiao

AU - Wang, Yaping

AU - Qin, Zengchang

AU - Lu, Jie

N1 - Publisher Copyright: © Springer International Publishing AG 2017.

PY - 2017

Y1 - 2017

N2 - We present a computerized image-based method to automatically identify aggressive tumors on combined positron emission tomography and magnetic resonance imaging (PET-MRI) using radiomics texture features from both PET and multi-parametric MRI (MP-MRI). The work aims at investigating the potential use of new composite textures from PET-MRI for the assessment of different biological properties present in cancer and non-cancer regions, and eventually for early detection of malignant tumors in real clinical practice. Towards this goal, a large number of radiomics features are extracted to characterize the intratumoural heterogeneity and microarchitectural morphologic differences within tumors. These image attributes are valuable for determining tumor aggressiveness. The radiomics model was evaluated on three types of cancers (pancreas, gallbladder, and liver). Compared to single image modality (PET or MRI), the fused PET and MP-MRI achieved the best classification performance in differentiating cancer and non-cancer regions with the area of under curve (AUC) of 0.87 for pancreas cancer, 0.89 for gallbladder cancer, and 0.82 for liver cancer. The results indicated that PET-MRI based imaging biomarkers could be useful in identifying aggressive tumors.

AB - We present a computerized image-based method to automatically identify aggressive tumors on combined positron emission tomography and magnetic resonance imaging (PET-MRI) using radiomics texture features from both PET and multi-parametric MRI (MP-MRI). The work aims at investigating the potential use of new composite textures from PET-MRI for the assessment of different biological properties present in cancer and non-cancer regions, and eventually for early detection of malignant tumors in real clinical practice. Towards this goal, a large number of radiomics features are extracted to characterize the intratumoural heterogeneity and microarchitectural morphologic differences within tumors. These image attributes are valuable for determining tumor aggressiveness. The radiomics model was evaluated on three types of cancers (pancreas, gallbladder, and liver). Compared to single image modality (PET or MRI), the fused PET and MP-MRI achieved the best classification performance in differentiating cancer and non-cancer regions with the area of under curve (AUC) of 0.87 for pancreas cancer, 0.89 for gallbladder cancer, and 0.82 for liver cancer. The results indicated that PET-MRI based imaging biomarkers could be useful in identifying aggressive tumors.

KW - Aggressive tumor

KW - Multi-parametric MRI

KW - PET

KW - Radiomics

UR - https://www.scopus.com/pages/publications/85035109625

U2 - 10.1007/978-3-319-70136-3_77

DO - 10.1007/978-3-319-70136-3_77

M3 - 会议稿件

AN - SCOPUS:85035109625

SN - 9783319701356

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

SP - 731

EP - 739

BT - Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings

A2 - Liu, Derong

A2 - Xie, Shengli

A2 - Zhao, Dongbin

A2 - Li, Yuanqing

A2 - El-Alfy, El-Sayed M.

PB - Springer Verlag

T2 - 24th International Conference on Neural Information Processing, ICONIP 2017

Y2 - 14 November 2017 through 18 November 2017

ER -

Wan T, Cui B, Wang Y, Qin Z, Lu J. A radiomics approach for automated identification of aggressive tumors on combined PET and multi-parametric MRI. In Liu D, Xie S, Zhao D, Li Y, El-Alfy ESM, editors, Neural Information Processing - 24th International Conference, ICONIP 2017, Proceedings. Springer Verlag. 2017. p. 731-739. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-70136-3_77

A radiomics approach for automated identification of aggressive tumors on combined PET and multi-parametric MRI

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this