Efficient machine learning models leveraging DCE-MRI morphological and dynamic features allow accurate breast lesion classification.

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: T2-weighted features: AUC 0.83; DCE dynamic features: AUC 0.91

• Primary Outcome: Area Under the Curve (AUC) for breast lesion classification

• Nnt Or Sensitivity: Sensitivity and specificity not reported

• Confidence Interval: T2: 95% CI 0.826-0.835; DCE: 95% CI 0.907-0.915

Clinical Application

Requires DCE-MRI capability and specialized radiomic analysis software Relevant to Australian breast screening programs; DCE-MRI available in major centres but may have PBS limitations for routine screening Women with suspicious breast lesions requiring MRI evaluation

Abstract

We propose an ensemble learning approach to classify malignant versus benign breast lesions by leveraging morphological and dynamic features derived from Magnetic Resonance Images (MRI). The analysis was performed on 164 breast lesions of the publicly available 'Advanced MRI Breast Lesions' dataset from The Cancer Imaging Archive, containing T2-weighted and Dynamic Contrast-Enhanced (DCE)-MRI sequences, along with the segmentation masks of suspicious lesions. After extracting radiomic features using the Pyradiomics Python package, we computed dynamic features from DCE-MRI kinetic curves, which describe the contrast agent wash-in and wash-out. These features have been defined as the derivatives of image intensity measures, such as mean and standard deviation, computed within the masks on the five DCE-MRI time steps. We trained and evaluated an eXtreme Gradient Boosting (XGBoost) classifier, experimenting with different combinations of features and segmentation masks in a stratified 5-fold cross-validation. The best model trained on T2-weighted MRI morphological features achieved an Area Under the Curve (AUC) score of 0.83 ± 0.04 (95% C.I. 0.826-0.835) on the independent test set consisting of 20 lesions, while the model using only dynamic features performed an AUC of 0.91 ± 0.03 (95% C.I. 0.907-0.915). Despite being obtained on a small test sample, these results show the potential of features derived from DCE images for breast lesions classification.