Overall Survival Time Prediction for High-grade Glioma Patients based on Large-scale Brain Functional Networks

Luyan Liu, Han Zhang, Jinsong Wu, Zhengda Yu, Xiaobo Chen, Islem Rekik, Qian Wang (Lead / Corresponding author), Junfeng Lu (Lead / Corresponding author), Dinggang Shen (Lead / Corresponding author)

Research output: Contribution to journalArticle

1 Citation (Scopus)
3 Downloads (Pure)

Abstract

High-grade glioma (HGG) is a lethal cancer with poor outcome. Accurate preoperative overall survival (OS) time prediction for HGG patients is crucial for treatment planning. Traditional presurgical and noninvasive OS prediction studies have used radiomics features at the local lesion area based on the magnetic resonance images (MRI). However, the highly complex lesion MRI appearance may have large individual variability, which could impede accurate individualized OS prediction. In this paper, we propose a novel concept, namely brain connectomics-based OS prediction. It is based on presurgical resting-state functional MRI (rs-fMRI) and the non-local, large-scale brain functional networks where the global and systemic prognostic features rather than the local lesion appearance are used to predict OS. We propose that the connectomics features could capture tumor-induced network-level alterations that are associated with prognosis. We construct both low-order (by means of sparse representation with regional rs-fMRI signals) and high-order functional connectivity (FC) networks (characterizing more complex multi-regional relationship by synchronized dynamics FC time courses). Then, we conduct a graph-theoretic analysis on both networks for a jointly, machine-learning-based individualized OS prediction. Based on a preliminary dataset (N = 34 with bad OS, mean OS, ~400 days; N = 34 with good OS, mean OS, ~1030 days), we achieve a promising OS prediction accuracy (86.8%) on separating the individuals with bad OS from those with good OS. However, if using only conventionally derived descriptive features (e.g., age and tumor characteristics), the accuracy is low (63.2%). Our study highlights the importance of the rs-fMRI and brain functional connectomics for treatment planning.
Original languageEnglish
Number of pages19
JournalBrain Imaging and Behavior
Early online date28 Aug 2018
DOIs
Publication statusE-pub ahead of print - 28 Aug 2018

Fingerprint

Brain
Magnetic resonance
Tumors
Planning
Learning systems
Magnetic Resonance Imaging

Keywords

  • Brain network
  • Connectomics
  • Functional connectivity
  • Glioma
  • Machine learning
  • Prognosis
  • Survival

Cite this

Liu, Luyan ; Zhang, Han ; Wu, Jinsong ; Yu, Zhengda ; Chen, Xiaobo ; Rekik, Islem ; Wang, Qian ; Lu, Junfeng ; Shen, Dinggang. / Overall Survival Time Prediction for High-grade Glioma Patients based on Large-scale Brain Functional Networks. In: Brain Imaging and Behavior. 2018.
@article{7167c8be2fc2469f8b55aa828a8fa0e1,
title = "Overall Survival Time Prediction for High-grade Glioma Patients based on Large-scale Brain Functional Networks",
abstract = "High-grade glioma (HGG) is a lethal cancer with poor outcome. Accurate preoperative overall survival (OS) time prediction for HGG patients is crucial for treatment planning. Traditional presurgical and noninvasive OS prediction studies have used radiomics features at the local lesion area based on the magnetic resonance images (MRI). However, the highly complex lesion MRI appearance may have large individual variability, which could impede accurate individualized OS prediction. In this paper, we propose a novel concept, namely brain connectomics-based OS prediction. It is based on presurgical resting-state functional MRI (rs-fMRI) and the non-local, large-scale brain functional networks where the global and systemic prognostic features rather than the local lesion appearance are used to predict OS. We propose that the connectomics features could capture tumor-induced network-level alterations that are associated with prognosis. We construct both low-order (by means of sparse representation with regional rs-fMRI signals) and high-order functional connectivity (FC) networks (characterizing more complex multi-regional relationship by synchronized dynamics FC time courses). Then, we conduct a graph-theoretic analysis on both networks for a jointly, machine-learning-based individualized OS prediction. Based on a preliminary dataset (N = 34 with bad OS, mean OS, ~400 days; N = 34 with good OS, mean OS, ~1030 days), we achieve a promising OS prediction accuracy (86.8{\%}) on separating the individuals with bad OS from those with good OS. However, if using only conventionally derived descriptive features (e.g., age and tumor characteristics), the accuracy is low (63.2{\%}). Our study highlights the importance of the rs-fMRI and brain functional connectomics for treatment planning.",
keywords = "Brain network, Connectomics, Functional connectivity, Glioma, Machine learning, Prognosis, Survival",
author = "Luyan Liu and Han Zhang and Jinsong Wu and Zhengda Yu and Xiaobo Chen and Islem Rekik and Qian Wang and Junfeng Lu and Dinggang Shen",
note = "This work is partially supported by National Natural Science Foundation of China (NSFC) Grants (61473190, 61401271, 81471733, 81201156, and 81401395), the National Key Technology R&D Program of China (2014BAI04B05), and NIH grant (EB022880).",
year = "2018",
month = "8",
day = "28",
doi = "10.1007/s11682-018-9949-2",
language = "English",
journal = "Brain Imaging and Behavior",
issn = "1931-7557",
publisher = "Springer",

}

Overall Survival Time Prediction for High-grade Glioma Patients based on Large-scale Brain Functional Networks. / Liu, Luyan ; Zhang, Han; Wu, Jinsong ; Yu, Zhengda ; Chen, Xiaobo ; Rekik, Islem; Wang, Qian (Lead / Corresponding author); Lu, Junfeng (Lead / Corresponding author); Shen, Dinggang (Lead / Corresponding author).

In: Brain Imaging and Behavior, 28.08.2018.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Overall Survival Time Prediction for High-grade Glioma Patients based on Large-scale Brain Functional Networks

AU - Liu, Luyan

AU - Zhang, Han

AU - Wu, Jinsong

AU - Yu, Zhengda

AU - Chen, Xiaobo

AU - Rekik, Islem

AU - Wang, Qian

AU - Lu, Junfeng

AU - Shen, Dinggang

N1 - This work is partially supported by National Natural Science Foundation of China (NSFC) Grants (61473190, 61401271, 81471733, 81201156, and 81401395), the National Key Technology R&D Program of China (2014BAI04B05), and NIH grant (EB022880).

PY - 2018/8/28

Y1 - 2018/8/28

N2 - High-grade glioma (HGG) is a lethal cancer with poor outcome. Accurate preoperative overall survival (OS) time prediction for HGG patients is crucial for treatment planning. Traditional presurgical and noninvasive OS prediction studies have used radiomics features at the local lesion area based on the magnetic resonance images (MRI). However, the highly complex lesion MRI appearance may have large individual variability, which could impede accurate individualized OS prediction. In this paper, we propose a novel concept, namely brain connectomics-based OS prediction. It is based on presurgical resting-state functional MRI (rs-fMRI) and the non-local, large-scale brain functional networks where the global and systemic prognostic features rather than the local lesion appearance are used to predict OS. We propose that the connectomics features could capture tumor-induced network-level alterations that are associated with prognosis. We construct both low-order (by means of sparse representation with regional rs-fMRI signals) and high-order functional connectivity (FC) networks (characterizing more complex multi-regional relationship by synchronized dynamics FC time courses). Then, we conduct a graph-theoretic analysis on both networks for a jointly, machine-learning-based individualized OS prediction. Based on a preliminary dataset (N = 34 with bad OS, mean OS, ~400 days; N = 34 with good OS, mean OS, ~1030 days), we achieve a promising OS prediction accuracy (86.8%) on separating the individuals with bad OS from those with good OS. However, if using only conventionally derived descriptive features (e.g., age and tumor characteristics), the accuracy is low (63.2%). Our study highlights the importance of the rs-fMRI and brain functional connectomics for treatment planning.

AB - High-grade glioma (HGG) is a lethal cancer with poor outcome. Accurate preoperative overall survival (OS) time prediction for HGG patients is crucial for treatment planning. Traditional presurgical and noninvasive OS prediction studies have used radiomics features at the local lesion area based on the magnetic resonance images (MRI). However, the highly complex lesion MRI appearance may have large individual variability, which could impede accurate individualized OS prediction. In this paper, we propose a novel concept, namely brain connectomics-based OS prediction. It is based on presurgical resting-state functional MRI (rs-fMRI) and the non-local, large-scale brain functional networks where the global and systemic prognostic features rather than the local lesion appearance are used to predict OS. We propose that the connectomics features could capture tumor-induced network-level alterations that are associated with prognosis. We construct both low-order (by means of sparse representation with regional rs-fMRI signals) and high-order functional connectivity (FC) networks (characterizing more complex multi-regional relationship by synchronized dynamics FC time courses). Then, we conduct a graph-theoretic analysis on both networks for a jointly, machine-learning-based individualized OS prediction. Based on a preliminary dataset (N = 34 with bad OS, mean OS, ~400 days; N = 34 with good OS, mean OS, ~1030 days), we achieve a promising OS prediction accuracy (86.8%) on separating the individuals with bad OS from those with good OS. However, if using only conventionally derived descriptive features (e.g., age and tumor characteristics), the accuracy is low (63.2%). Our study highlights the importance of the rs-fMRI and brain functional connectomics for treatment planning.

KW - Brain network

KW - Connectomics

KW - Functional connectivity

KW - Glioma

KW - Machine learning

KW - Prognosis

KW - Survival

UR - http://www.scopus.com/inward/record.url?scp=85053237709&partnerID=8YFLogxK

U2 - 10.1007/s11682-018-9949-2

DO - 10.1007/s11682-018-9949-2

M3 - Article

JO - Brain Imaging and Behavior

JF - Brain Imaging and Behavior

SN - 1931-7557

ER -