Online citations, reference lists, and bibliographies.
← Back to Search

Predicting Breast Cancer Molecular Subtype With MRI Dataset Utilizing Convolutional Neural Network Algorithm

R. Ha, Simukayi Mutasa, Jenika Karcich, N. Gupta, E. Sant, J. Nemer, M. Sun, P. Chang, M. Liu, S. Jambawalikar
Published 2019 · Computer Science, Medicine

Cite This
Download PDF
Analyze on Scholarcy
Share
To develop a convolutional neural network (CNN) algorithm that can predict the molecular subtype of a breast cancer based on MRI features. An IRB-approved study was performed in 216 patients with available pre-treatment MRIs and immunohistochemical staining pathology data. First post-contrast MRI images were used for 3D segmentation using 3D slicer. A CNN architecture was designed with 14 layers. Residual connections were used in the earlier layers to allow stabilization of gradients during backpropagation. Inception style layers were utilized deeper in the network to allow learned segregation of more complex feature mappings. Extensive regularization was utilized including dropout, L2, feature map dropout, and transition layers. The class imbalance was addressed by doubling the input of underrepresented classes and utilizing a class sensitive cost function. Parameters were tuned based on a 20% validation group. A class balanced holdout set of 40 patients was utilized as the testing set. Software code was written in Python using the TensorFlow module on a Linux workstation with one NVidia Titan X GPU. Seventy-four luminal A, 106 luminal B, 13 HER2+, and 23 basal breast tumors were evaluated. Testing set accuracy was measured at 70%. The class normalized macro area under receiver operating curve (ROC) was measured at 0.853. Non-normalized micro-aggregated AUC was measured at 0.871, representing improved discriminatory power for the highly represented Luminal A and Luminal B subtypes. Aggregate sensitivity and specificity was measured at 0.603 and 0.958. MRI analysis of breast cancers utilizing a novel CNN can predict the molecular subtype of breast cancers. Larger data sets will likely improve our model.
This paper references
Gradient methods for minimizing composite objective function
Y. Nesterov (2007)
10.1142/S1793351X16500045
Deep Learning
X. Hao (2016)
10.1109/5.726791
Gradient-based learning applied to document recognition
Y. LeCun (1998)
10.1016/S0140-6736(11)61350-0
MRI for breast cancer screening, diagnosis, and treatment
M. Morrow (2011)
10.1073/pnas.191367098
Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications
T. Sørlie (2001)
10.1007/s12282-013-0512-0
Intratumoral heterogeneity of the distribution of kinetic parameters in breast cancer: comparison based on the molecular subtypes of invasive breast cancer
K. Yamaguchi (2014)
10.1109/ICCV.2017.97
Revisiting Unreasonable Effectiveness of Data in Deep Learning Era
C. Sun (2017)
10.3322/caac.21208
Cancer statistics, 2014
R. Siegel (2014)
10.1093/annonc/mdr304
Strategies for subtypes—dealing with the diversity of breast cancer: highlights of the St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2011
A. Goldhirsch (2011)
10.1200/JCO.2012.46.1574
Patterns of Recurrence and outcome according to breast cancer subtypes in lymph node-negative disease: results from international breast cancer study group trials VIII and IX.
Otto Metzger-Filho (2013)
10.1016/J.BREASTDIS.2015.10.008
Identification of Intrinsic Imaging Phenotypes for Breast Cancer Tumors: Preliminary Associations with Gene Expression Profiles
K. Shin (2015)
10.1148/radiol.14132641
Radiogenomic analysis of breast cancer: luminal B molecular subtype is associated with enhancement dynamics at MR imaging.
M. Mazurowski (2014)
10.1148/radiol.14121031
Computerized image analysis for identifying triple-negative breast cancers and differentiating them from other molecular subtypes of breast cancer on dynamic contrast-enhanced MR images: a feasibility study.
S. Agner (2014)
10.1109/cvpr.2016.90
Deep Residual Learning for Image Recognition
Kaiming He (2016)
10.1016/j.mric.2009.09.005
Diagnostic breast MR imaging: current status and future directions.
E. Morris (2010)
Rectified Linear Units Improve Restricted Boltzmann Machines
V. Nair (2010)
10.1245/s10434-009-0606-2
Presenting Features of Breast Cancer Differ by Molecular Subtype
L. Wiechmann (2009)
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
S. Ioffe (2015)
10.2214/AJR.180.4.1800901
MR imaging of the ipsilateral breast in women with percutaneously proven breast cancer.
L. Liberman (2003)
10.1186/s12880-017-0239-z
MRI texture analysis in differentiating luminal A and luminal B breast cancer molecular subtypes - a feasibility study
Kirsi Holli-Helenius (2017)
10.1117/1.JMI.2.4.041007
Prediction of clinical phenotypes in invasive breast carcinomas from the integration of radiomics and genomics data
Wentian Guo (2015)
10.1148/radiol.15142698
Breast Cancer: Radiogenomic Biomarker Reveals Associations among Dynamic Contrast-enhanced MR Imaging, Long Noncoding RNA, and Metastasis.
S. Yamamoto (2015)
10.1148/radiol.13132195
Behind the numbers: Decoding molecular phenotypes with radiogenomics--guiding principles and technical considerations.
M. Kuo (2014)
10.1148/radiol.09090838
Cancerous breast lesions on dynamic contrast-enhanced MR images: computerized characterization for image-based prognostic markers.
N. Bhooshan (2010)
10.1002/jmri.24879
Computational approach to radiogenomics of breast cancer: Luminal A and luminal B molecular subtypes are associated with imaging features on routine breast MRI extracted using computer vision algorithms
L. Grimm (2015)
Incorporating Nesterov Momentum into Adam
Timothy Dozat (2016)
Adam: A Method for Stochastic Optimization
Diederik P. Kingma (2015)
10.1200/JCO.2007.14.4287
Breast cancer subtype approximated by estrogen receptor, progesterone receptor, and HER-2 is associated with local and distant recurrence after breast-conserving therapy.
P. Nguyen (2008)
10.1118/1.1695652
Computerized interpretation of breast MRI: investigation of enhancement-variance dynamics.
W. Chen (2004)
10.1001/JAMA.295.21.2492
Race, breast cancer subtypes, and survival in the Carolina Breast Cancer Study.
L. Carey (2006)
10.1016/j.mri.2012.06.010
Radiomics: the process and the challenges.
Virendra Kumar (2012)
10.1371/journal.pone.0171683
Radiomic analysis reveals DCE-MRI features for prediction of molecular subtypes of breast cancer
Ming Fan (2017)
10.1016/J.EJSO.2004.02.003
Contrast-enhanced MR imaging of breast lesions and effect on treatment.
K. Schelfout (2004)
10.1093/annonc/mds586
Molecular subclasses of breast cancer: how do we define them? The IMPAKT 2012 Working Group Statement.
S. Guiu (2012)
10.1158/1078-0432.CCR-06-1109
The Triple Negative Paradox: Primary Tumor Chemosensitivity of Breast Cancer Subtypes
L. Carey (2007)
10.1117/12.2295471
Breast cancer molecular subtype classification using deep features: preliminary results
Z. Zhu (2018)
Understanding the difficulty of training deep feedforward neural networks
Xavier Glorot (2010)
: comparison based on the molecular subtypes of invasive breast cancer
E Blaschke (2015)
10.1088/0031-9155/56/18/014
Computerized three-class classification of MRI-based prognostic markers for breast cancer.
N. Bhooshan (2011)
10.1038/nature14539
Deep Learning
I. Goodfellow (2015)
10.2214/AJR.14.13666
Breast cancer molecular subtype as a predictor of the utility of preoperative MRI.
R. Ha (2015)
10.1002/jmri.24884
MRI phenotype of breast cancer: Kinetic assessment for molecular subtypes
Eric M Blaschke (2015)
Dropout: a simple way to prevent neural networks from overfitting
Nitish Srivastava (2014)
10.1038/35021093
Molecular portraits of human breast tumours
C. Perou (2000)



This paper is referenced by
10.1101/2020.07.22.215558
Identifying transcription patterns of histology and radiomics features in NSCLC with neural networks
Nova F. Smedley (2020)
10.3390/app9173589
MLW-gcForest: A Multi-Weighted gcForest Model for Cancer Subtype Classification by Methylation Data
Yunyun Dong (2019)
10.1016/j.clinimag.2020.04.025
Understanding artificial intelligence based radiology studies: What is overfitting?
Simukayi Mutasa (2020)
10.1007/978-3-030-59830-3_56
An Integrated Deep Architecture for Lesion Detection in Breast MRI
Ghazal Rouhafzay (2020)
10.1093/bioinformatics/btaa126
Discovering and interpreting transcriptomic drivers of imaging traits using neural networks
Nova F. Smedley (2020)
10.1016/j.kint.2020.02.027
Artificial intelligence and machine learning in nephropathology.
Jan U. Becker (2020)
10.1155/2020/6805710
Current Status and Future Perspectives of Artificial Intelligence in Magnetic Resonance Breast Imaging
A. Meyer-Bäse (2020)
10.1016/j.mri.2020.08.021
A novel CNN algorithm for pathological complete response prediction using an I-SPY TRIAL breast MRI database.
M. Liu (2020)
Deep learning-based prediction of response to HER2-targeted neoadjuvant chemotherapy from pre-treatment dynamic breast MRI: A multi-institutional validation study
Nathaniel Braman (2020)
10.1007/s00330-020-07274-x
Prediction of breast cancer molecular subtypes on DCE-MRI using convolutional neural network with transfer learning between two centers
Y. Zhang (2020)
10.3892/ijo.2020.5063
Artificial intelligence radiogenomics for advancing precision and effectiveness in oncologic care (Review)
Eleftherios Trivizakis (2020)
10.1007/s00330-020-07277-8
Contrast-enhanced cone beam breast CT features of breast cancers: correlation with immunohistochemical receptors and molecular subtypes
Yue Ma (2020)
10.1155/2019/6978650
Molecular Subtypes Recognition of Breast Cancer in Dynamic Contrast-Enhanced Breast Magnetic Resonance Imaging Phenotypes from Radiomics Data
Wei Li (2019)
10.1016/j.eswa.2020.113501
Deep learning radiomics in breast cancer with different modalities: Overview and future
Ting Pang (2020)
10.1002/jmri.26852
Machine learning in breast MRI
B. Reig (2019)
10.1016/j.diii.2019.02.008
Detection and characterization of MRI breast lesions using deep learning.
P. Hérent (2019)
10.1016/J.EJRAD.2019.03.015
Prediction of molecular subtypes of breast cancer using BI-RADS features based on a "white box" machine learning approach in a multi-modal imaging setting.
Mingxiang Wu (2019)
Semantic Scholar Logo Some data provided by SemanticScholar