Abstract:
Breast cancer is a complex and heterogeneous disease with varying clinical outcomes and treatment responses among different subtypes. Accurate classification of breast cancer subtypes is crucial for personalized treatment planning. This study presents an pplication of Radiomics, a non-invasive approach that extracts quantitative features from medical images, for predicting breast cancer subtypes from 3D MRI scans. The study utilized a dataset of breast MRI images from a diverse group of patients diagnosed with various breast cancer subtypes. A comprehensive set of radiomic features were extracted using Pyradiomics. A machine learning pipeline integrating feature selection and classification algorithms was developed. The predictive model was evaluated using performance metrics. The results demonstrate the promising performance of the proposed approach in accurately predicting breast cancer subtypes. Moreover, the study highlights the potential clinical utility of non-invasive radiomics-based subtype prediction and potentially improves patient outcomes by helping doctors make informed decisions.