Could you please provide the script or exact pipeline to convert any DICOM volume into the tensors you provided for the challenge (Brain extracted 240*240*155 isotropic resampled volumes or any other additional issues you covered.) I know there is a detailed pipeline description in your arxiv paper. However minute differences may lead to big deflections. I want to check the generalization ability of my trained model on the volumes obtained in our hospitals with my colleagues prosepectively. Thank you in advance