Related Experiment Video
Updated: Jul 28, 2026
Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging
Published on: December 15, 2014
Radiomics analysis of intratumoral and different peritumoral regions from multiparametric MRI for evaluating HER2
Jing Zhou1, Xuan Yu1, Qingxia Wu2
1Department of Medical Imaging, Henan Provincial People's Hospital & People's Hospital of Zhengzhou University & Imaging Diagnosis of Neurological Diseases and Research Laboratory of Henan Province, Zhengzhou, 450003, Henan Province, China.
Radiomics signatures from multiparametric MRI can noninvasively predict HER2 status in breast cancer. Combining intratumoral and peritumoral regions improves accuracy, with the optimal signature achieving an AUC of 0.795.
Area of Science:
- Radiology
- Oncology
- Medical Imaging
Background:
- Accurate HER2 status assessment is crucial for breast cancer treatment decisions.
- Multiparametric MRI offers a noninvasive approach to evaluate tumor characteristics.
Purpose of the Study:
- To investigate the potential of radiomics signatures (RSs) from intratumoral and peritumoral regions on multiparametric MRI to noninvasively evaluate HER2 status in breast cancer.
Main Methods:
- Retrospective analysis of 992 breast cancer patients undergoing preoperative MRI.
- Extraction of quantitative radiomics features from DCE-T1, T2, and DWI sequences.
- Construction of RSs using SVM for HER2 status prediction, exploring various region of interest (ROI) combinations.
Main Results:
- Single-region RSs achieved AUCs up to 0.719.
- Combined intratumoral and peritumoral RSs improved prediction, with AUCs reaching 0.754.
- The optimal RS combining three sequences (DCE-T1, T2, DWI) across multiple ROIs achieved an AUC of 0.795.
Conclusions:
- Radiomics signatures derived from multiparametric MRI can effectively predict HER2 status in breast cancer.
- Incorporating both intratumoral and peritumoral information enhances predictive performance.
- The study identified an optimal radiomics signature for noninvasive HER2 status evaluation.
