Effect of different molecular subtype reference standards in AI training: implications for DCE-MRI radiomics of breast cancers

A single breast cancer lesion can have different luminal molecular subtyping when using either immunohistochemical (IHC) staining alone or the St. Gallen criteria that includes Ki-67. This may impact artificial intelligence/computer aided diagnosis (AI/CADx) for determining molecular subtype from medical images. We investigated this using 28 radiomic features extracted from DCE-MR images of 877 unique lesions segmented by a fuzzy c-means method, for three groups of lesions: (1) Luminal A lesions by both reference standards (“agreement”), (2) lesions that were Luminal A by IHC and Luminal B by St. Gallen (“disagreement”), and (3) Luminal B lesions by both reference standards (“agreement”). The Kruskal-Wallis (KW) test for statistically significant differences in groups of lesions was sequentially followed by the Mann-Whitney U test to determine pair-wise statistical difference between groups for relevant features from the KW test. Classification of lesions as Luminal A or Luminal B using all available radiomic features was conducted using three sets of lesions: (1) lesions with IHC alone molecular subtyping, (2) lesions with St. Gallen molecular subtyping, and (3) agreement lesions. Five-fold cross-validation using stepwise feature selection/linear discriminant analysis classifier classified lesions in each set, with performance measured by the area under the receiver operating characteristic curve (AUC). Six features (sphericity, irregularity, surface area to volume ratio, variance of radial gradient histogram, sum average, and volume of most enhancing voxels) were significantly different among the three groups of features with mixed difference of the disagreement group of lesions to the two agreement luminal groups. When using agreement lesions, more features were selected for classification and the AUC was significantly higher (P < 0.003) than using lesions subtyped by either reference standard. The results suggest that the disagreement of reference standards may impact the development of medical imaging AI/CADx methods for determining molecular subtype.

[1]  Teoria Statistica Delle Classi e Calcolo Delle Probabilità , 2022, The SAGE Encyclopedia of Research Design.

[2]  K. Ng,et al.  Magnetic Resonance Imaging Phenotypes of Breast Cancer Molecular Subtypes: A Systematic Review. , 2021, Academic radiology.

[3]  M. Giger,et al.  Robustness of radiomic features of benign breast lesions and hormone receptor positive/HER2-negative cancers across DCE-MR magnet strengths. , 2021, Magnetic resonance imaging.

[4]  M. Giger,et al.  Improved Classification of Benign and Malignant Breast Lesions Using Deep Feature Maximum Intensity Projection MRI in Breast Cancer Diagnosis Using Dynamic Contrast-enhanced MRI. , 2021, Radiology. Artificial intelligence.

[5]  K. Czene,et al.  Concordance of Immunohistochemistry-Based and Gene Expression-Based Subtyping in Breast Cancer , 2020, JNCI cancer spectrum.

[6]  Karen Drukker,et al.  Case-based repeatability of machine learning classification performance on breast MRI , 2020, Medical Imaging.

[7]  Karen Drukker,et al.  Repeatability profiles towards consistent sensitivity and specificity levels for machine learning on breast DCE-MRI , 2020, Medical Imaging.

[8]  Maryellen L. Giger,et al.  Harmonization of radiomic features of breast lesions across international DCE-MRI datasets , 2020, Journal of medical imaging.

[9]  Mingying Du,et al.  Differentiation between Luminal A and B Molecular Subtypes of Breast Cancer Using Pharmacokinetic Quantitative Parameters with Histogram and Texture Features on Preoperative Dynamic Contrast-Enhanced Magnetic Resonance Imaging. , 2020, Academic radiology.

[10]  M. Giger,et al.  Independent validation of machine learning in diagnosing breast Cancer on magnetic resonance imaging within a single institution , 2019, Cancer Imaging.

[11]  Danny F. Martinez,et al.  Radiomic signatures with contrast-enhanced magnetic resonance imaging for the assessment of breast cancer receptor status and molecular subtypes: initial results , 2019, Breast Cancer Research.

[12]  L. Saal,et al.  Agreement between molecular subtyping and surrogate subtype classification: a contemporary population-based study of ER-positive/HER2-negative primary breast cancer , 2019, Breast Cancer Research and Treatment.

[13]  Maryellen L. Giger,et al.  Additive Benefit of Radiomics Over Size Alone in the Distinction Between Benign Lesions and Luminal A Cancers on a Large Clinical Breast MRI Dataset. , 2019, Academic radiology.

[14]  Maryellen L. Giger,et al.  Effect of biopsy on the MRI radiomics classification of benign lesions and luminal A cancers , 2018, Other Conferences.

[15]  Ruijiang Li,et al.  Identifying relations between imaging phenotypes and molecular subtypes of breast cancer: Model discovery and external validation , 2017, Journal of magnetic resonance imaging : JMRI.

[16]  Lihua Li,et al.  Radiomic analysis reveals DCE-MRI features for prediction of molecular subtypes of breast cancer , 2017, PloS one.

[17]  K. Lång,et al.  ESTIMATES OF BREAST CANCER GROWTH RATE FROM MAMMOGRAMS AND ITS RELATION TO TUMOUR CHARACTERISTICS. , 2016, Radiation protection dosimetry.

[18]  Erich P Huang,et al.  Quantitative MRI radiomics in the prediction of molecular classifications of breast cancer subtypes in the TCGA/TCIA data set , 2016, npj Breast Cancer.

[19]  Lars J. Grimm,et al.  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 , 2015, Journal of magnetic resonance imaging : JMRI.

[20]  R. Gelber,et al.  Tailoring therapies—improving the management of early breast cancer: St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2015 , 2015, Annals of oncology : official journal of the European Society for Medical Oncology.

[21]  S Michiels,et al.  Molecular subclasses of breast cancer: how do we define them? The IMPAKT 2012 Working Group Statement. , 2012, Annals of oncology : official journal of the European Society for Medical Oncology.

[22]  R. Gelber,et al.  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 , 2011, Annals of oncology : official journal of the European Society for Medical Oncology.

[23]  G. Colditz,et al.  Comparison of molecular phenotypes of ductal carcinoma in situ and invasive breast cancer , 2008, Breast Cancer Research.

[24]  M. Giger,et al.  Volumetric texture analysis of breast lesions on contrast‐enhanced magnetic resonance images , 2007, Magnetic resonance in medicine.

[25]  M. Giger,et al.  Automatic identification and classification of characteristic kinetic curves of breast lesions on DCE-MRI. , 2006, Medical physics.

[26]  M. Giger,et al.  A fuzzy c-means (FCM)-based approach for computerized segmentation of breast lesions in dynamic contrast-enhanced MR images. , 2006, Academic radiology.

[27]  C. Metz,et al.  "Proper" Binormal ROC Curves: Theory and Maximum-Likelihood Estimation. , 1999, Journal of mathematical psychology.

[28]  M L Giger,et al.  Computerized analysis of breast lesions in three dimensions using dynamic magnetic-resonance imaging. , 1998, Medical physics.