As our team was working on Glioblastoma RANO rating classification on the LUMIERE dataset, we were firstly looking at the Pyradiomics features based on DeepBraTumIA segmentations provided as a csv file in the figshare website. (https://figshare.com/articles/dataset/LUMIERE_dataset_-_Pyradiomics_features_based_on_DeepBraTumIA_segmentations/21187033?file=37681080) We understand that features were extracted from the images resampled to atlas space. Also, If I understand correctly, then these features are extracted from different types of tumor segmentations (necrosis, edema, contrast enhancing tumor). However, we had a few questions regarding the feature extraction process for this file. 1. Were these features extracted before or after intensity normalization was performed on the original images? 2. If these features were extracted after intensity normalization, then how exactly was the intensity normalization performed? Was it subject based or cohort based? We hope to get a response regarding these questions soon as it would help expedite our research on the LUMIERE data set.