Efficient Segmentation of Breast Cancer CT Images: A Multi-Threshold Image Segmentation Method Based on Learning Search Algorithm

Early diagnosis of breast cancer is crucial for improving patient survival rates, with medical imaging technology playing a key role. Traditional image segmentation methods suffer from limitations such as sensitivity to thresholds and a tendency to become trapped in local optima. Metaheuristic algorithms (MAs) offer an effective strategy to address these issues. This study introduces a novel optimization algorithm—Learning Search Algorithm (LSA)—that simulates human learning behaviors, including historical experience, social role modeling, and proactive learning, to optimize multi-threshold image segmentation. The LSA employs dynamic adaptive control parameters during the global search and local exploitation phases, enhancing segmentation accuracy and convergence speed. In the application of breast cancer CT image segmentation, the LSA outperforms existing advanced algorithms, providing higher quality solutions. This research not only enriches the repertoire of image segmentation algorithms but also provides new technical support for the early diagnosis and treatment of breast cancer.