Improving CNN-Based Breast Cancer Detection Integrating Quantum Layers

Abstract

Breast cancer continues to be a significant burden on global healthcare systems, as early and accurate diagnosis is crucial for improving patient outcomes. Conventional methods used for diagnosis include mammography and biopsy; although they do supply critical information, they often have poor accuracy and are operator-dependent. Artificial Intelligence(AI), particularly Convolutional Neural Networks, presents a promising tool for analyzing medical images; however, conventional CNNs face significant challenges in generalizing from one dataset to another. This paper presents a hybrid Quantum Convolutional Neural Networks(QCNN) framework by integrating the classical feature extraction models VGG16, VGG19, and InceptionV3 with a Quantum Convolutional Layer (QCL). It uses the principles of quantum, such as superposition and entanglement, which process high-dimensional data for capturing non-linear patterns. Therefore, it improves the model's accuracy, sensitivity, and specificity. This hybrid framework presents a scalable and robust solution for the early detection of breast cancer, thereby advancing automated diagnostic systems to enhance reliability and adaptability. © 2025 IEEE.