Breast Cancer Detection in Computed Tomography Images Using Attention-Based Convolutional Neural Networks

In this paper, propose a model for detecting breast cancer using an Attention Convolutional Neural Network (ACNN) integrated with a specialized detection module that undergoes rigorous training and testing. To optimize the cancer detection process, the model has been pretrained to incorporate all essential features necessary for accurate diagnosis. The images obtained from computed tomography (CT) scans are pre-processed before being fed into the model for classification, ensuring optimal results during testing. To assess the model's reliability in identifying cancer, we employ a 5 -fold cross-validation technique. Experimental validation is performed using Python simulations, which enable thorough evaluation of the model's performance. The results indicate that the proposed ACNN model demonstrates robust capabilities in detecting cancer across large and complex image datasets. When compared to existing methods, the simulation outcomes show that our model significantly reduces detection errors and improves the accuracy of cancer identification. Overall, our study highlights the potential of Attention-based Convolutional Neural Networks in providing a highly reliable approach to cancer detection, making it a valuable tool for medical image analysis. The model's ability to minimize detection errors marks a significant advancement over traditional methods, offering new insights into the efficient and accurate identification of cancerous lesions in CT images.