No abstract

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 microaggregated 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.

Recent innovations in quantitative magnetic resonance imaging (MRI) measurement methods have led to improvements in accuracy, repeatability, and acquisition speed, and have prompted renewed interest to reevaluate the medical value of quantitative T1. The purpose of this study was to determine the bias and reproducibility of T1 measurements in a variety of MRI systems with an eye toward assessing the feasibility of applying diagnostic threshold T1 measurement across multiple clinical sites. We used the International Society of Magnetic Resonance in Medicine/National Institute of Standards and Technology (ISMRM/NIST) system phantom to assess variations of T1 measurements, using a slow, reference standard inversion recovery sequence and a rapid, commonly-available variable flip angle sequence, across MRI systems at 1.5 tesla (T) (two vendors, with number of MRI systems n = 9) and 3 T (three vendors, n = 18). We compared the T1 measurements from inversion recovery and variable flip angle scans to ISMRM/NIST phantom reference values using Analysis of Variance (ANOVA) to test for statistical differences between T1 measurements grouped according to MRI scanner manufacturers and/or static field strengths. The inversion recovery method had minor over- and under-estimations compared to the NMR-measured T1 values at both 1.5 T and 3 T. Variable flip angle measurements had substantially greater deviations from the NMR-measured T1 values than the inversion recovery measurements. At 3 T, the measured variable flip angle T1 for one vendor is significantly different than the other two vendors for most of the samples throughout the clinically relevant range of T1. There was no consistent pattern of discrepancy between vendors. We suggest establishing rigorous quality control procedures for validating quantitative MRI methods to promote confidence and stability in associated measurement techniques and to enable translation of diagnostic threshold from the research center to the entire clinical community.

No abstract

2 Technical Efficacy: Stage2 J. MAGN. RESON. IMAGING 2017;46:690-696.