Exploring Kinetic Curves Features for the Classification of Benign and Malignant Breast Lesions in DCE-MRI

TL;DR

This study tackles differentiating benign and malignant breast lesions in DCE-MRI by introducing a fully automated 3D framework that fuses dynamic kinetic-curve features with radiomic features derived from the entire tumor volume. The pipeline segments the tumor with a joint-phase attention (JPA) network, extracts 11 dynamic features from voxel-wise kinetics ($IERM$/$DERM$ maps) and 800 radiomic features from original and filtered images, selects 68 features via LASSO (10 dynamic, 58 radiomic), and classifies with Linear Discriminant Analysis on the combined feature set. On an in-house dataset of 200 scans (298 tumors), the combined features achieve the best performance with an AUC of $0.9476$, clearly outperforming dynamic-only ($AUC=0.8253$) or radiomic-only ($AUC=0.9080$) approaches. The results demonstrate the value of integrating 3D kinetic curves with radiomics for more accurate, automated breast lesion classification, and point to future improvements in segmentation and classifier choices, including deep learning-based end-to-end approaches.

Abstract

Breast cancer is the most common malignant tumor among women and the second cause of cancer-related death. Early diagnosis in clinical practice is crucial for timely treatment and prognosis. Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) has revealed great usability in the preoperative diagnosis and assessing therapy effects thanks to its capability to reflect the morphology and dynamic characteristics of breast lesions. However, most existing computer-assisted diagnosis algorithms only consider conventional radiomic features when classifying benign and malignant lesions in DCE-MRI. In this study, we propose to fully leverage the dynamic characteristics from the kinetic curves as well as the radiomic features to boost the classification accuracy of benign and malignant breast lesions. The proposed method is a fully automated solution by directly analyzing the 3D features from the DCE-MRI. The proposed method is evaluated on an in-house dataset including 200 DCE-MRI scans with 298 breast tumors (172 benign and 126 malignant tumors), achieving favorable classification accuracy with an area under curve (AUC) of 0.94. By simultaneously considering the dynamic and radiomic features, it is beneficial to effectively distinguish between benign and malignant breast lesions. The algorithm is publicly available at https://github.com/ryandok/JPA.

Figures (8)

• Figure 1: The proposed 3D framework for the classification of benign and malignant breast tumors in DCE-MRI.
• Figure 2: Different types of the kinetic curves.
• Figure 3: DCE-MRI sequences and the segmented tumor regions.
• Figure 4: The selected dynamic features and their corresponding coefficients on the LASSO model.
• Figure 5: The selected radiomic features and their corresponding coefficients on the LASSO model.