Multi-label inductive matrix completion for joint MGMT and IDH1 status prediction for glioma patients

Lei Chen, Han Zhang, Kim Han Thung, Luyan Liu, Junfeng Lu, Jinsong Wu, Qian Wang, Dinggang Shen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

MGMT promoter methylation and IDH1 mutation in high-grade gliomas (HGG) have proven to be the two important molecular indicators associated with better prognosis. Traditionally, the statuses of MGMT and IDH1 are obtained via surgical biopsy, which is laborious, invasive and time-consuming. Accurate presurgical prediction of their statuses based on preoperative imaging data is of great clinical value towards better treatment plan. In this paper, we propose a novel Multi-label Inductive Matrix Completion (MIMC) model, highlighted by the online inductive learning strategy, to jointly predict both MGMT and IDH1 statuses. Our MIMC model not only uses the training subjects with possibly missing MGMT/IDH1 labels, but also leverages the unlabeled testing subjects as a supplement to the limited training dataset. More importantly, we learn inductive labels, instead of directly using transductive labels, as the prediction results for the testing subjects, to alleviate the overfitting issue in small-sample-size studies. Furthermore, we design an optimization algorithm with guaranteed convergence based on the block coordinate descent method to solve the multivariate non-smooth MIMC model. Finally, by using a precious single-center multi-modality presurgical brain imaging and genetic dataset of primary HGG, we demonstrate that our method can produce accurate prediction results, outperforming the previous widely-used single- or multi-task machine learning methods. This study shows the promise of utilizing imaging-derived brain connectome phenotypes for prognosis of HGG in a non-invasive manner.

Original languageEnglish
Title of host publicationMedical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings
PublisherSpringer Verlag
Pages450-458
Number of pages9
Volume10434 LNCS
ISBN (Print)9783319661841
DOIs
Publication statusPublished - 2017
Externally publishedYes
Event20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017 - Quebec City, Canada
Duration: 2017 Sep 112017 Sep 13

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10434 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017
CountryCanada
CityQuebec City
Period17/9/1117/9/13

Fingerprint

Matrix Completion
Labels
Prognosis
Imaging
Prediction
Multi-task Learning
Inductive Learning
Coordinate Descent
Descent Method
Multimodality
Testing
Learning Strategies
Online Learning
Overfitting
Small Sample Size
Imaging techniques
Promoter
Phenotype
Leverage
Brain

Keywords

  • High-grade gliomas
  • Matrix completion
  • Molecular biomarker

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Chen, L., Zhang, H., Thung, K. H., Liu, L., Lu, J., Wu, J., ... Shen, D. (2017). Multi-label inductive matrix completion for joint MGMT and IDH1 status prediction for glioma patients. In Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings (Vol. 10434 LNCS, pp. 450-458). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10434 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-66185-8_51

Multi-label inductive matrix completion for joint MGMT and IDH1 status prediction for glioma patients. / Chen, Lei; Zhang, Han; Thung, Kim Han; Liu, Luyan; Lu, Junfeng; Wu, Jinsong; Wang, Qian; Shen, Dinggang.

Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. Vol. 10434 LNCS Springer Verlag, 2017. p. 450-458 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10434 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Chen, L, Zhang, H, Thung, KH, Liu, L, Lu, J, Wu, J, Wang, Q & Shen, D 2017, Multi-label inductive matrix completion for joint MGMT and IDH1 status prediction for glioma patients. in Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. vol. 10434 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10434 LNCS, Springer Verlag, pp. 450-458, 20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017, Quebec City, Canada, 17/9/11. https://doi.org/10.1007/978-3-319-66185-8_51
Chen L, Zhang H, Thung KH, Liu L, Lu J, Wu J et al. Multi-label inductive matrix completion for joint MGMT and IDH1 status prediction for glioma patients. In Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. Vol. 10434 LNCS. Springer Verlag. 2017. p. 450-458. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-66185-8_51
Chen, Lei ; Zhang, Han ; Thung, Kim Han ; Liu, Luyan ; Lu, Junfeng ; Wu, Jinsong ; Wang, Qian ; Shen, Dinggang. / Multi-label inductive matrix completion for joint MGMT and IDH1 status prediction for glioma patients. Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. Vol. 10434 LNCS Springer Verlag, 2017. pp. 450-458 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{d6ad4c7af27c40038f1e925c2ab74107,
title = "Multi-label inductive matrix completion for joint MGMT and IDH1 status prediction for glioma patients",
abstract = "MGMT promoter methylation and IDH1 mutation in high-grade gliomas (HGG) have proven to be the two important molecular indicators associated with better prognosis. Traditionally, the statuses of MGMT and IDH1 are obtained via surgical biopsy, which is laborious, invasive and time-consuming. Accurate presurgical prediction of their statuses based on preoperative imaging data is of great clinical value towards better treatment plan. In this paper, we propose a novel Multi-label Inductive Matrix Completion (MIMC) model, highlighted by the online inductive learning strategy, to jointly predict both MGMT and IDH1 statuses. Our MIMC model not only uses the training subjects with possibly missing MGMT/IDH1 labels, but also leverages the unlabeled testing subjects as a supplement to the limited training dataset. More importantly, we learn inductive labels, instead of directly using transductive labels, as the prediction results for the testing subjects, to alleviate the overfitting issue in small-sample-size studies. Furthermore, we design an optimization algorithm with guaranteed convergence based on the block coordinate descent method to solve the multivariate non-smooth MIMC model. Finally, by using a precious single-center multi-modality presurgical brain imaging and genetic dataset of primary HGG, we demonstrate that our method can produce accurate prediction results, outperforming the previous widely-used single- or multi-task machine learning methods. This study shows the promise of utilizing imaging-derived brain connectome phenotypes for prognosis of HGG in a non-invasive manner.",
keywords = "High-grade gliomas, Matrix completion, Molecular biomarker",
author = "Lei Chen and Han Zhang and Thung, {Kim Han} and Luyan Liu and Junfeng Lu and Jinsong Wu and Qian Wang and Dinggang Shen",
year = "2017",
doi = "10.1007/978-3-319-66185-8_51",
language = "English",
isbn = "9783319661841",
volume = "10434 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "450--458",
booktitle = "Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings",

}

TY - GEN

T1 - Multi-label inductive matrix completion for joint MGMT and IDH1 status prediction for glioma patients

AU - Chen, Lei

AU - Zhang, Han

AU - Thung, Kim Han

AU - Liu, Luyan

AU - Lu, Junfeng

AU - Wu, Jinsong

AU - Wang, Qian

AU - Shen, Dinggang

PY - 2017

Y1 - 2017

N2 - MGMT promoter methylation and IDH1 mutation in high-grade gliomas (HGG) have proven to be the two important molecular indicators associated with better prognosis. Traditionally, the statuses of MGMT and IDH1 are obtained via surgical biopsy, which is laborious, invasive and time-consuming. Accurate presurgical prediction of their statuses based on preoperative imaging data is of great clinical value towards better treatment plan. In this paper, we propose a novel Multi-label Inductive Matrix Completion (MIMC) model, highlighted by the online inductive learning strategy, to jointly predict both MGMT and IDH1 statuses. Our MIMC model not only uses the training subjects with possibly missing MGMT/IDH1 labels, but also leverages the unlabeled testing subjects as a supplement to the limited training dataset. More importantly, we learn inductive labels, instead of directly using transductive labels, as the prediction results for the testing subjects, to alleviate the overfitting issue in small-sample-size studies. Furthermore, we design an optimization algorithm with guaranteed convergence based on the block coordinate descent method to solve the multivariate non-smooth MIMC model. Finally, by using a precious single-center multi-modality presurgical brain imaging and genetic dataset of primary HGG, we demonstrate that our method can produce accurate prediction results, outperforming the previous widely-used single- or multi-task machine learning methods. This study shows the promise of utilizing imaging-derived brain connectome phenotypes for prognosis of HGG in a non-invasive manner.

AB - MGMT promoter methylation and IDH1 mutation in high-grade gliomas (HGG) have proven to be the two important molecular indicators associated with better prognosis. Traditionally, the statuses of MGMT and IDH1 are obtained via surgical biopsy, which is laborious, invasive and time-consuming. Accurate presurgical prediction of their statuses based on preoperative imaging data is of great clinical value towards better treatment plan. In this paper, we propose a novel Multi-label Inductive Matrix Completion (MIMC) model, highlighted by the online inductive learning strategy, to jointly predict both MGMT and IDH1 statuses. Our MIMC model not only uses the training subjects with possibly missing MGMT/IDH1 labels, but also leverages the unlabeled testing subjects as a supplement to the limited training dataset. More importantly, we learn inductive labels, instead of directly using transductive labels, as the prediction results for the testing subjects, to alleviate the overfitting issue in small-sample-size studies. Furthermore, we design an optimization algorithm with guaranteed convergence based on the block coordinate descent method to solve the multivariate non-smooth MIMC model. Finally, by using a precious single-center multi-modality presurgical brain imaging and genetic dataset of primary HGG, we demonstrate that our method can produce accurate prediction results, outperforming the previous widely-used single- or multi-task machine learning methods. This study shows the promise of utilizing imaging-derived brain connectome phenotypes for prognosis of HGG in a non-invasive manner.

KW - High-grade gliomas

KW - Matrix completion

KW - Molecular biomarker

UR - http://www.scopus.com/inward/record.url?scp=85029543610&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029543610&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-66185-8_51

DO - 10.1007/978-3-319-66185-8_51

M3 - Conference contribution

AN - SCOPUS:85029543610

SN - 9783319661841

VL - 10434 LNCS

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

SP - 450

EP - 458

BT - Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings

PB - Springer Verlag

ER -