Automated quantitative assessment of amorphous calcifications: Towards improved malignancy risk stratification

Abstract

Background: Amorphous calcifications noted on mammograms (i.e., small and indistinct calcifications that are difficult to characterize) are associated with high diagnostic uncertainty, often leading to biopsies. Yet, only 20% of biopsied amorphous calcifications are cancer. We present a quantitative approach for distinguishing between benign and actionable (high-risk and malignant) amorphous calcifications using a combination of local textures, global spatial relationships, and interpretable handcrafted expert features.

Method: Our approach was trained and validated on a set of 168 2D full-field digital mammography exams (248 images) from 168 patients. Within these 248 images, we identified 276 image regions with segmented amorphous calcifications and a biopsy-confirmed diagnosis. A set of local (radiomic and region measurements) and global features (distribution and expert-defined) were extracted from each image. Local features were grouped using an unsupervised k-means clustering algorithm. All global features were concatenated with clustered local features and used to train a LightGBM classifier to distinguish benign from actionable cases.

Results: On the held-out test set of 60 images, our approach achieved a sensitivity of 100%, specificity of 35%, and a positive predictive value of 38% when the decision threshold was set to 0.4. Given that all of the images in our test set resulted in a recommendation of a biopsy, the use of our algorithm would have identified 15 images (25%) that were benign, potentially reducing the number of breast biopsies.

Conclusions: Quantitative analysis of full-field digital mammograms can extract subtle shape, texture, and distribution features that may help to distinguish between benign and actionable amorphous calcifications.

Keywords: Machine learning; Mammography; Microcalcifications; Radiomics.

Fig. 1.

Cohort selection.

Fig. 2.

Classification pipeline.

Fig. 3.

(a) An example ROI, (b) the foreground mask, and (c) the background mask.

Fig. 4.

Visualizations of dilated foreground masks at two representative scales.

Fig. 5.

K means clustering-based feature aggregation pipeline. During training, we generated clusters using the features extracted from the objects of training ROIs. During testing, we utilized the clusters created during the training phase to predict the clusters of the objects from the testing ROIs. The process was repeated to generate three local feature sets followed by their concatenation with global features.

Fig. 6.

ROC curve of the classification using (a) K-means clustering-based aggregation of local textural features and global features with LightGBM classifier (b) Features extracted from fine-tuned VGG-16 using weighted cross-entropy loss + LightGBM classifier (c) Features extracted from fine-tuned ResNet-50 using weighted cross-entropy loss + LightGBM classifier (d) Mean and standard deviation aggregation of local features + LightGBM classifier.