A machine learning approach to discriminate malignant and benign breast lesions using multimodal MRI

The study explores using Artificial Intelligence to distinguish malignant from benign breast lesions with multimodal MRI. Using a public dataset of 200 patients, a radiomic approach has been tried to extract the features to train an XGBoost classifier, alongside the extraction of dynamic features from the kinetic curves of DCE-MRI. The model achieved an AUC of 0.90 using radiomic features alone, 0.91 using dynamic features and 0.92 when combining radiomic and dynamic features, showing the potential diagnostic improvement with both modalities