Value of Radiomics Based on DCE-MRI in distinguishing benign and malignant breast lesions: Predicting Histological Grade and Lymph Node Metastasis of Breast Cancer

Abstract

Objective: The characteristics of benign and malignant breast lesions often overlap and intersect, leading to missed diagnosis or inaccurate diagnosis and excessive biopsy, besides surgical procedures. This study aimed to assess the diagnostic value of radiomics based on dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) in evaluating the pathological characteristics of breast cancer.

Methodology: This retrospective case-control study included 110 patients with breast lesions who underwent DCE-MRI and obtained pathological results at Yongkang First People’s Hospital and Yongkang Hospital of Traditional Chinese Medicine from September 2019 to December 2024. According to the results, 55 patients with confirmed breast cancer lesions (malignant group) were matched with a cohort of patients with benign lesions (benign group) at a 1:1 ratio. Radiomic parameters from DCE-MRI were analyzed in the two groups.

Results: The radiomic features derived from DCE-MRI showed significant differences not only between benign and malignant breast lesions but also among subgroups stratified by histological grade and axillary lymph node metastasis. Specifically, patients with higher histological grades demonstrated elevated values in long run emphasis and all-angle cluster prominence variance (×1013), and decreased values in all-angle correlation, surface-to-volume ratio, and uniformity. Similarly, patients with axillary lymph node metastasis exhibited significantly higher long-run emphasis and cluster prominence variance, and lower uniformity and all-angle correlation (all P < 0.05).

Conclusion: DCE-MRI-based radiomics can effectively differentiate benign and malignant breast lesions and has a predictive value for assessing histological grade and axillary lymph node status. Radiomic features may therefore serve as noninvasive imaging biomarkers to support breast cancer diagnosis, grading, and staging.