Abstract

Breast cancer’s heterogeneity poses significant challenges in its diagnosis and treatment, necessitating advancements in non-invasive diagnostic methods. Radiomics, an emerging field that leverages mathematical analysis of medical imaging, offers the potential for designing personalized treatment strategies in breast cancer care. This systematic review aimed to synthesize current research on the application of radiomics in breast cancer diagnosis and treatment assessment, with a focus on magnetic resonance imaging (MRI). Following the preferred reporting items for systematic reviews and meta-analyses extension for systematic reviews protocol, a comprehensive literature search was conducted across major databases and grey literature. The review included peer-reviewed articles published in 2015 onward, focusing exclusively on studies involving MRI imaging in radiomics. The studies were categorized into four themes including tumor subtype classification, prediction of pathologically complete response, detection of lymph node metastasis, and prediction of recurrence. The review included 48 articles, revealing significant insights into the use of radiomics for breast cancer analysis. These studies highlighted the superiority of MRI, particularly with 3T scanners and multiparametric protocols in radiomic analysis. The findings underscore the potential of combining radiomics with clinical data for enhanced prediction accuracy in tumor characterization, treatment response, and recurrence risk. The review also illuminated the importance of intratumor heterogeneity, as well as the challenges associated with image acquisition, segmentation, and feature extraction in radiomic studies. Radiomics in MRI shows considerable promise in advancing breast cancer diagnosis and treatment, aligning with the goals of personalized medicine.

Key words: Radiomics, breast cancer, MRI, personalized medicine, systematic review