Residual convolutional neural network for the determination of IDH status in low- and high-grade gliomas from mr imaging

Ken Chang, Harrison X. Bai, Hao Zhou, Chang Su, Wenya Linda Bi, Ena Agbodza, Vasileios K. Kavouridis, Joeky T. Senders, Alessandro Boaro, Andrew Beers, Biqi Zhang, Alexandra Capellini, Weihua Liao, Qin Shen, Xuejun Li, Bo Xiao, Jane Cryan, Shakti Ramkissoon, Lori Ramkissoon, Keith Ligon & 10 others Patrick Y. Wen, Ranjit S. Bindra, John Woo, Omar Arnaout, Elizabeth R. Gerstner, Paul J. Zhang, Bruce R. Rosen, Li Yang, Raymond Y. Huang, Jayashree Kalpathy-Cramer

Research output: Contribution to journalArticle

38 Citations (Scopus)

Abstract

Purpose: Isocitrate dehydrogenase (IDH) mutations in glioma patients confer longer survival and may guide treatment decision making. We aimed to predict the IDH status of gliomas from MR imaging by applying a residual convolutional neural network to preoperative radiographic data. Experimental Design: Preoperative imaging was acquired for 201 patients from the Hospital of University of Pennsylvania (HUP), 157 patients from Brigham and Women's Hospital (BWH), and 138 patients from The Cancer Imaging Archive (TCIA) and divided into training, validation, and testing sets. We trained a residual convolutional neural network for each MR sequence (FLAIR, T2, T1 precontrast, and T1 postcontrast) and built a predictive model from the outputs. To increase the size of the training set and prevent overfitting, we augmented the training set images by introducing random rotations, translations, flips, shearing, and zooming. Results: With our neural network model, we achieved IDH prediction accuracies of 82.8% (AUC ¼ 0.90), 83.0% (AUC ¼ 0.93), and 85.7% (AUC ¼ 0.94) within training, validation, and testing sets, respectively. When age at diagnosis was incorporated into the model, the training, validation, and testing accuracies increased to 87.3% (AUC ¼ 0.93), 87.6% (AUC ¼ 0.95), and 89.1% (AUC ¼ 0.95), respectively. Conclusions: We developed a deep learning technique to noninvasively predict IDH genotype in grade II–IV glioma using conventional MR imaging using a multi-institutional data set.

Original languageEnglish (US)
Pages (from-to)1073-1081
Number of pages9
JournalClinical Cancer Research
Volume24
Issue number5
DOIs
StatePublished - Mar 1 2018
Externally publishedYes

Fingerprint

Isocitrate Dehydrogenase
Glioma
Area Under Curve
Neural Networks (Computer)
Decision Making
Research Design
Genotype
Learning
Mutation
Survival
Neoplasms

ASJC Scopus subject areas

  • Oncology
  • Cancer Research

Cite this

Residual convolutional neural network for the determination of IDH status in low- and high-grade gliomas from mr imaging. / Chang, Ken; Bai, Harrison X.; Zhou, Hao; Su, Chang; Bi, Wenya Linda; Agbodza, Ena; Kavouridis, Vasileios K.; Senders, Joeky T.; Boaro, Alessandro; Beers, Andrew; Zhang, Biqi; Capellini, Alexandra; Liao, Weihua; Shen, Qin; Li, Xuejun; Xiao, Bo; Cryan, Jane; Ramkissoon, Shakti; Ramkissoon, Lori; Ligon, Keith; Wen, Patrick Y.; Bindra, Ranjit S.; Woo, John; Arnaout, Omar; Gerstner, Elizabeth R.; Zhang, Paul J.; Rosen, Bruce R.; Yang, Li; Huang, Raymond Y.; Kalpathy-Cramer, Jayashree.

In: Clinical Cancer Research, Vol. 24, No. 5, 01.03.2018, p. 1073-1081.

Research output: Contribution to journalArticle

Chang, K, Bai, HX, Zhou, H, Su, C, Bi, WL, Agbodza, E, Kavouridis, VK, Senders, JT, Boaro, A, Beers, A, Zhang, B, Capellini, A, Liao, W, Shen, Q, Li, X, Xiao, B, Cryan, J, Ramkissoon, S, Ramkissoon, L, Ligon, K, Wen, PY, Bindra, RS, Woo, J, Arnaout, O, Gerstner, ER, Zhang, PJ, Rosen, BR, Yang, L, Huang, RY & Kalpathy-Cramer, J 2018, 'Residual convolutional neural network for the determination of IDH status in low- and high-grade gliomas from mr imaging', Clinical Cancer Research, vol. 24, no. 5, pp. 1073-1081. https://doi.org/10.1158/1078-0432.CCR-17-2236
Chang, Ken ; Bai, Harrison X. ; Zhou, Hao ; Su, Chang ; Bi, Wenya Linda ; Agbodza, Ena ; Kavouridis, Vasileios K. ; Senders, Joeky T. ; Boaro, Alessandro ; Beers, Andrew ; Zhang, Biqi ; Capellini, Alexandra ; Liao, Weihua ; Shen, Qin ; Li, Xuejun ; Xiao, Bo ; Cryan, Jane ; Ramkissoon, Shakti ; Ramkissoon, Lori ; Ligon, Keith ; Wen, Patrick Y. ; Bindra, Ranjit S. ; Woo, John ; Arnaout, Omar ; Gerstner, Elizabeth R. ; Zhang, Paul J. ; Rosen, Bruce R. ; Yang, Li ; Huang, Raymond Y. ; Kalpathy-Cramer, Jayashree. / Residual convolutional neural network for the determination of IDH status in low- and high-grade gliomas from mr imaging. In: Clinical Cancer Research. 2018 ; Vol. 24, No. 5. pp. 1073-1081.
@article{9f358ba1391542ed8ea34366a8d4e2d1,
title = "Residual convolutional neural network for the determination of IDH status in low- and high-grade gliomas from mr imaging",
abstract = "Purpose: Isocitrate dehydrogenase (IDH) mutations in glioma patients confer longer survival and may guide treatment decision making. We aimed to predict the IDH status of gliomas from MR imaging by applying a residual convolutional neural network to preoperative radiographic data. Experimental Design: Preoperative imaging was acquired for 201 patients from the Hospital of University of Pennsylvania (HUP), 157 patients from Brigham and Women's Hospital (BWH), and 138 patients from The Cancer Imaging Archive (TCIA) and divided into training, validation, and testing sets. We trained a residual convolutional neural network for each MR sequence (FLAIR, T2, T1 precontrast, and T1 postcontrast) and built a predictive model from the outputs. To increase the size of the training set and prevent overfitting, we augmented the training set images by introducing random rotations, translations, flips, shearing, and zooming. Results: With our neural network model, we achieved IDH prediction accuracies of 82.8{\%} (AUC ¼ 0.90), 83.0{\%} (AUC ¼ 0.93), and 85.7{\%} (AUC ¼ 0.94) within training, validation, and testing sets, respectively. When age at diagnosis was incorporated into the model, the training, validation, and testing accuracies increased to 87.3{\%} (AUC ¼ 0.93), 87.6{\%} (AUC ¼ 0.95), and 89.1{\%} (AUC ¼ 0.95), respectively. Conclusions: We developed a deep learning technique to noninvasively predict IDH genotype in grade II–IV glioma using conventional MR imaging using a multi-institutional data set.",
author = "Ken Chang and Bai, {Harrison X.} and Hao Zhou and Chang Su and Bi, {Wenya Linda} and Ena Agbodza and Kavouridis, {Vasileios K.} and Senders, {Joeky T.} and Alessandro Boaro and Andrew Beers and Biqi Zhang and Alexandra Capellini and Weihua Liao and Qin Shen and Xuejun Li and Bo Xiao and Jane Cryan and Shakti Ramkissoon and Lori Ramkissoon and Keith Ligon and Wen, {Patrick Y.} and Bindra, {Ranjit S.} and John Woo and Omar Arnaout and Gerstner, {Elizabeth R.} and Zhang, {Paul J.} and Rosen, {Bruce R.} and Li Yang and Huang, {Raymond Y.} and Jayashree Kalpathy-Cramer",
year = "2018",
month = "3",
day = "1",
doi = "10.1158/1078-0432.CCR-17-2236",
language = "English (US)",
volume = "24",
pages = "1073--1081",
journal = "Clinical Cancer Research",
issn = "1078-0432",
publisher = "American Association for Cancer Research Inc.",
number = "5",

}

TY - JOUR

T1 - Residual convolutional neural network for the determination of IDH status in low- and high-grade gliomas from mr imaging

AU - Chang, Ken

AU - Bai, Harrison X.

AU - Zhou, Hao

AU - Su, Chang

AU - Bi, Wenya Linda

AU - Agbodza, Ena

AU - Kavouridis, Vasileios K.

AU - Senders, Joeky T.

AU - Boaro, Alessandro

AU - Beers, Andrew

AU - Zhang, Biqi

AU - Capellini, Alexandra

AU - Liao, Weihua

AU - Shen, Qin

AU - Li, Xuejun

AU - Xiao, Bo

AU - Cryan, Jane

AU - Ramkissoon, Shakti

AU - Ramkissoon, Lori

AU - Ligon, Keith

AU - Wen, Patrick Y.

AU - Bindra, Ranjit S.

AU - Woo, John

AU - Arnaout, Omar

AU - Gerstner, Elizabeth R.

AU - Zhang, Paul J.

AU - Rosen, Bruce R.

AU - Yang, Li

AU - Huang, Raymond Y.

AU - Kalpathy-Cramer, Jayashree

PY - 2018/3/1

Y1 - 2018/3/1

N2 - Purpose: Isocitrate dehydrogenase (IDH) mutations in glioma patients confer longer survival and may guide treatment decision making. We aimed to predict the IDH status of gliomas from MR imaging by applying a residual convolutional neural network to preoperative radiographic data. Experimental Design: Preoperative imaging was acquired for 201 patients from the Hospital of University of Pennsylvania (HUP), 157 patients from Brigham and Women's Hospital (BWH), and 138 patients from The Cancer Imaging Archive (TCIA) and divided into training, validation, and testing sets. We trained a residual convolutional neural network for each MR sequence (FLAIR, T2, T1 precontrast, and T1 postcontrast) and built a predictive model from the outputs. To increase the size of the training set and prevent overfitting, we augmented the training set images by introducing random rotations, translations, flips, shearing, and zooming. Results: With our neural network model, we achieved IDH prediction accuracies of 82.8% (AUC ¼ 0.90), 83.0% (AUC ¼ 0.93), and 85.7% (AUC ¼ 0.94) within training, validation, and testing sets, respectively. When age at diagnosis was incorporated into the model, the training, validation, and testing accuracies increased to 87.3% (AUC ¼ 0.93), 87.6% (AUC ¼ 0.95), and 89.1% (AUC ¼ 0.95), respectively. Conclusions: We developed a deep learning technique to noninvasively predict IDH genotype in grade II–IV glioma using conventional MR imaging using a multi-institutional data set.

AB - Purpose: Isocitrate dehydrogenase (IDH) mutations in glioma patients confer longer survival and may guide treatment decision making. We aimed to predict the IDH status of gliomas from MR imaging by applying a residual convolutional neural network to preoperative radiographic data. Experimental Design: Preoperative imaging was acquired for 201 patients from the Hospital of University of Pennsylvania (HUP), 157 patients from Brigham and Women's Hospital (BWH), and 138 patients from The Cancer Imaging Archive (TCIA) and divided into training, validation, and testing sets. We trained a residual convolutional neural network for each MR sequence (FLAIR, T2, T1 precontrast, and T1 postcontrast) and built a predictive model from the outputs. To increase the size of the training set and prevent overfitting, we augmented the training set images by introducing random rotations, translations, flips, shearing, and zooming. Results: With our neural network model, we achieved IDH prediction accuracies of 82.8% (AUC ¼ 0.90), 83.0% (AUC ¼ 0.93), and 85.7% (AUC ¼ 0.94) within training, validation, and testing sets, respectively. When age at diagnosis was incorporated into the model, the training, validation, and testing accuracies increased to 87.3% (AUC ¼ 0.93), 87.6% (AUC ¼ 0.95), and 89.1% (AUC ¼ 0.95), respectively. Conclusions: We developed a deep learning technique to noninvasively predict IDH genotype in grade II–IV glioma using conventional MR imaging using a multi-institutional data set.

UR - http://www.scopus.com/inward/record.url?scp=85047769155&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85047769155&partnerID=8YFLogxK

U2 - 10.1158/1078-0432.CCR-17-2236

DO - 10.1158/1078-0432.CCR-17-2236

M3 - Article

VL - 24

SP - 1073

EP - 1081

JO - Clinical Cancer Research

JF - Clinical Cancer Research

SN - 1078-0432

IS - 5

ER -