Brain Tumor Survival Prediction Using Pseudo-Labeling

Created using ChatSlide

This presentation explores the classification and survival prediction of gliomas using MRI, emphasizing the challenges in prediction due to independent networks and censored data. A novel End-to-End model framework is proposed, incorporating pseudo-labeling and enhanced datasets. The discussion includes the architectural details of U-Net and the benefits of pseudo-labeling, supported by comparative analysis results. Concluding with the model's contributions, it suggests future directions for...