Breast cancer segmentation then detection using TransUNet

I am looking for help with a project related to breast cancer segmentation using TransUNet and classification after segementation . The goal is to develop a system that can segment breast cancer tissue and then accurately detect it. As part of the project, data augmentation (including random crops, flipping, and rotation) will be used to bolster the accuracy of the network. Finally, the core of the system will be a TransUNet architecture. If you are experienced in developing this type of system, I would love to hear from you. USe the break his break cancer dataset https://www.kaggle.com/datasets/ambarish/breakhis