Deep Learning Based Early Diagnostic System for Metastatic Breast Cancer Mutation Using MATLAB

In this work, we proposed a deep learning- based diagnostic system leveraging convolutional neural networks (CNNs) to analyse histopathological images for early detection of metastatic breast cancer with their gene mutation and expression level. The dataset utilized in this study comprises images obtained from the Genomic Data Commons (GDC) portal, accompanied by pertinent patient information including case-id, gender, primary site, primary diagnosis, race, disease type, ethnicity, gene mutation and expression level. A simple CNN architecture was employed for its ability to autonomously learn discriminative features from input images. The model yielded accuracy of 96%. Additionally, we have implemented a streamlined retrieval mechanism enabling seamless access to specific dataset indices along with corresponding image samples and associated information, facilitating clinical interpretation. The proposed system demonstrates promising results in the early diagnosis of metastatic breast cancer, offering potential utility in clinical settings for improved patient outcomes.