Detection and Classification of Invasive Ductal Carcinoma using Artificial Intelligence

Abstract One of the most prevalent causes of death for women is breast cancer. An essential component of treating invasive ductal carcinoma (IDC) is early identification and detection. The accuracy of diagnosis can be greatly increased by computer-aided methods. Artificial intelligence (AI) has the potential to be used in breast imaging for a variety of activities, increasing beyond the current application in computer-aided diagnosis to include detection, risk assessment, response to therapy, prognosis, and classification. To recognize IDC cells, high-resolution microscope histopathology images are used by pathologists. But this approach is time-consuming, tedious, and arbitrary because of the slight variations between the IDC and Not IDC. The automatic detection of IDC in histopathology images has been established using a new algorithm based on artificial intelligence (AI) intelligence to address these issues. Furthermore, an adaptive Mask Region-based Convolutional Network (Mask R-CNN) and Deep CNNs are described for multistage IDC-cell detection. The model is simulated and evaluated by extensively evaluating the proposed breast cancer histopathology image dataset. The results of the experiments demonstrate that the proposed approach can significantly improve performance. The proposed model for IDC classification achieved a maximum accuracy of 99.02%, precision of 98.93%, recall of 98.35%, and an F1 score of 99.23%. Extensive experiments and comparisons with baseline methods demonstrate the effectiveness of the proposed approach.