In the model-spinal-rootlets project, I used a nnUNet model to predict a complete BIDS dataset.
To ensure reproducibility, I have created a script that extracts all images of a given contrast from the BIDS dataset and move/copy them to a designated folder.
I also added an --suffix argument to allow users to convert the image names from BIDS format to nnUNet format if needed.
Perhaps this script can be integrated into the @naga-karthik script run_nnunet_inference.py.
Unless you think that extracting images from a BIDS dataset may have uses beyond nnUNet prediction.
In the model-spinal-rootlets project, I used a nnUNet model to predict a complete BIDS dataset. To ensure reproducibility, I have created a script that extracts all images of a given contrast from the BIDS dataset and move/copy them to a designated folder.
I also added an --suffix argument to allow users to convert the image names from BIDS format to nnUNet format if needed.
Perhaps this script can be integrated into the @naga-karthik script run_nnunet_inference.py. Unless you think that extracting images from a BIDS dataset may have uses beyond nnUNet prediction.