International Journal of Science, Engineering and TechnologyAuthors: Lakshmi Shree A, J Tanuja, Prerana K N, V Satish
Abstract: Accurate visualization of breast cancer tumors is critical for effective diagnosis and treatment planning. However, conventional 2D medical imaging often lacks the spatial depth needed to fully assess tumor morphology. This research presents a deep learning-based approach for 3D reconstruction of breast cancer tumors using 2D medical scans. A U-Net architecture is employed to segment tumors from MRI or CT images, followed by a 3D Generative Adversarial Network (3D-GAN) to reconstruct volumetric tumor models. The final 3D structures are rendered using the Visualization Toolkit (VTK), enabling interactive exploration of tumor size, shape, and localization. The system enhances diagnostic accuracy, supports clinical workflows, and provides a scalable framework for future applications in other cancer types. Experimental results demonstrate the model’s effectiveness in generating anatomically consistent 3D reconstructions from limited 2D data.
DOI:
