Augmented Reality-Assisted Breast Mass Segmentation Using Hybrid Fuzzy-Deformable Model

Breast cancer is one of the leading causes of death among women, making early and accurate diagnosis crucial. Computer-aided diagnosis (CAD) serves as a valuable tool to assist radiologists in detecting and analyzing breast masses. This research develops a CAD system for detecting masses in the mini-MIAS dataset, aiming for optimal segmentation to support physicians in decision-making, especially before surgery. The proposed approach first applies anisotropic filtering for noise suppression and contrast enhancement using mathematical morphology. Then, mass segmentation is performed using a hybrid method combining Cooperative Fuzzy Possibilistic techniques with a Deformable Model (Level Set). Experimental results show significant improvements in mass extraction. Additionally, augmented reality (AR) is explored to enhance breast cancer detection by overlaying segmented tumor regions onto real-world images via AR-enabled devices. This approach provides an intuitive visualization, improving diagnostic accuracy, surgical planning, and patient understanding.