Breast Cancer Detection and Classification Using Adaptively Regularized Fuzzy C-Means Based on Kernels Inception Transformer Quantum Generative Adversarial Network with Emperor Penguin Optimization

Accurately identifying the specific type of breast tumor (BT) is significant for developing a higher efficient treatment strategy. Even now, one of the main causes of mortality for women worldwide is BT. This paper presents an advanced framework for the diagnosis and classification of BT using the inception transformer quantum generative adversarial network optimized by Emperor Penguin Optimization (IT-QGAN-EPO). The system preprocesses BreakHis and BUSI images using the Difference of Gaussian Filtering for noise reduction and clarity enhancement. Precise segmentation is achieved using the Adapted Regulated Kernel-Based Fuzzy C-Means (ARKFCM) algorithm, ensuring accurate identification of breast regions. Feature extraction and classification leverage the IT-QGAN architecture, optimized for efficient learning and superior performance. The proposed framework achieves a remarkable 99.9% accuracy for and 99.8% recall for the Break His dataset and 98.7% accuracy and 98.9% recall for the BUSI dataset, outperforming existing methodologies. This approach significantly enhances the detection of complex tumor structures, reduces false positives, and demonstrates scalability for large-scale medical datasets, making it an effective tool for clinical applications in BT diagnosis.