Classification of Background Parenchymal Uptake on Molecular Breast Imaging Using a Convolutional Neural Network

Abstract

PURPOSE Background parenchymal uptake (BPU), which describes the level of radiotracer uptake in normal fibroglandular tissue on molecular breast imaging (MBI), has been identified as a breast cancer risk factor. Our objective was to develop and validate a deep learning model using image convolution to automatically categorize BPU on MBI. METHODS MBI examinations obtained for clinical and research purposes from 2004 to 2015 were reviewed to classify the BPU pattern using a standardized five-category scale. Two expert radiologists provided interpretations that were used as the reference standard for modeling. The modeling consisted of training and validating a convolutional neural network to predict BPU. Model performance was summarized in data reserved to test the performance of the algorithm at the per-image and per-breast levels. RESULTS Training was performed on 24,639 images from 3,133 unique patients. The model performance on the withheld testing data (6,172 images; 786 patients) was evaluated. Using direct matching on the predicted classification resulted in an accuracy of 69.4% (95% CI, 67.4% to 71.3%), and if prediction within one category was considered, accuracy increased to 96.0% (95% CI, 95.2% to 96.7%). When considering the breast-level prediction of BPU, the accuracy remained strong, with 70.3% (95% CI, 68.0% to 72.6%) and 96.2% (95% CI, 95.3% to 97.2%) for the direct match and allowance for one category, respectively. CONCLUSION BPU provided a robust target for training a convolutional neural network. A validated computer algorithm will allow for objective, reproducible encoding of BPU to foster its integration into risk-stratification algorithms.

Monte Carlo‐derived 99mTc uptake quantification with commercial planar MBI: Tumor and breast activity concentrations

“…12 While BPU is typically qualitatively assessed by an expert radiologist using a 5-level scale, 13 quantitative BPU techniques have been investigated using mammogram contour registration 14 or convolutional neural networks. 15 Our work demonstrates the feasibility of extending these BPU measurements to also include absolute normal tissue uptake estimates. Additional investigations will be needed to first properly implement this technique in patient images, where parenchymal uptake is not homogeneous and background ROIs throughout the same breast can differ in average pixel counts by up to 35%, 6 and to then determine correlation, if any, between absolute normal tissue activity concentrations to BPU classes and breast cancer risk.…”

“…reported that increased background parenchymal uptake (BPU), defined as increased tracer uptake in fibroglandular tissue relative to uptake in subcutaneous fat, was an independent risk factor for breast cancer 12 . While BPU is typically qualitatively assessed by an expert radiologist using a 5‐level scale, 13 quantitative BPU techniques have been investigated using mammogram contour registration 14 or convolutional neural networks 15 . Our work demonstrates the feasibility of extending these BPU measurements to also include absolute normal tissue uptake estimates.…”

Artificial Intelligence in Cardiology: Present and Future

“…Although there are many types of neural networks, broadly speaking, deep neural networks permit sophisticated recognition of subtle patterns in a nonlinear manner using models that contain many layers of data abstraction and synthesisdresulting in an uncanny ability to "read" mammograms and electrocardiograms (ECGs) or recognize faces. 5 Although deep neural networks do in fact provide deep intelligence, at present it is a narrow intelligence with very focused skills and limited spontaneous adaptability or general intelligence.…”