A classification model of breast masses in DCE-MRI using kinetic curves features with quantum-Raina’s polynomial based fusion

Abstract

Listen

Translate article

Breast cancer is the most common cancer among women worldwide and is currently the second most common cancer-related death in women. However, early detection and diagnosis of breast cancer can lead to a complete remission and can extend survival periods. Dynamic Contrast Enhanced MRI (DCE-MRI) is being increasingly used in early detection and characterization of breast lesions. It has higher accuracy in breast cancer detection than other imaging modalities currently utilized in breast imaging as it offers better insight into tumor morphology and microenvironment. The most important problem is that earlier research did not consider breast cancer characteristics’ features that might be useful for accurately identifying breast malignancy and differentiating it from the more common benign breast pathologies. The use of recently developed image feature extraction algorithms in conjunction with radiological imaging helps in the diagnosis and categorization of breast cancer. With this view, the current study presents a novel fusion model using the kinetic curves features with a proposed quantum-Raina’s polynomial features for the breast cancer classification in MRI. This study uses two approaches for features extractions of DCE-MRI scans of the breast, namely, kinetic and QRP features.The excellent performance of the fusion extracted features improves the classification accuracy of breast into benign and malignant lesions. The maximum achieved accuracy for classifying a dataset comprising 300 DCE-MRI breast scans is 97.4%. The accuracy of the proposed classification model is significant with a low complexity rate.