Challenges and Future Directions in Breast Cancer Segmentation: A Research Perspective
Year: 2025
Type: article
Abstract: Breast cancer is one of the most aggressive and widespread illnesses afflicting women across the globe. Fast and precise segmentation techniques are essential for early detection, diagnosis, and treatment planning. This paper reviews comprehensively segmentation methods used in breast cancer detection operations, including traditional methods of thresholding and edge detection and eliciting advanced deep learning techniques such as Convolutional Neural Networks (CNN), U-Net, Generative Adversarial Networks (GANs), and Transformer-based models. The review stresses the merits of hybrid approaches combining many segmentation paradigms for better accuracy and robustness. This new round of research underlines the recent progress in segmentation with the help of attention mechanisms, precise mapping, and multimodal imaging integration. Yet, problems such as dataset-level issues, generalization issues, computational complexity, and lack of explainability still remain. Future research will design lightweight architectures, explainable AI, federated learning, and advanced multimodal data fusion techniques. This paper highlights the dynamic nature of breast cancer segmentation and marks that without continued innovation, achieving clinically relevant and accurate automated segmentation systems will remain a challenge.
Authors: Swathi Iyer, Salwa Tudilkar, Srivaramangai R
Institution: University of Mumbai
Language: English
Topic: AI in cancer detection
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4407693752