You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Allow CSV files as input to the normalization CLIs, where there is an image (or several modalities) column to be normalized, a mask column, an output column, etc. This is mostly relevant for sample-based methods (RAVEL, Nyul, and LSQ).
Motivation
The current workflow for sample-based methods requires images to be setup in a directory structure which is often unnatural for how neuroimaging/medical image projects are formatted. A CSV file specifying all the image paths would be an alternative that also makes sure the masks are in proper correspondence with the images.
The CSV file format could also be used to normalize multiple modalities sequentially, e.g., with FCM you could provide the CSV with the T1-w, T2-w, etc. image paths under corresponding columns and the method can automatically find the tissue mask on the T1-w image and use the tissue mask to normalize the other modalities.
Alternatives
Force users to symlink/copy their images into a directory to normalize them.
The text was updated successfully, but these errors were encountered:
🚀 Feature
Allow CSV files as input to the normalization CLIs, where there is an image (or several modalities) column to be normalized, a mask column, an output column, etc. This is mostly relevant for sample-based methods (RAVEL, Nyul, and LSQ).
Motivation
The current workflow for sample-based methods requires images to be setup in a directory structure which is often unnatural for how neuroimaging/medical image projects are formatted. A CSV file specifying all the image paths would be an alternative that also makes sure the masks are in proper correspondence with the images.
The CSV file format could also be used to normalize multiple modalities sequentially, e.g., with FCM you could provide the CSV with the T1-w, T2-w, etc. image paths under corresponding columns and the method can automatically find the tissue mask on the T1-w image and use the tissue mask to normalize the other modalities.
Alternatives
Force users to symlink/copy their images into a directory to normalize them.
The text was updated successfully, but these errors were encountered: