Links

Direct data upload

The Direct Upload functionality allows users to upload their image data directly to RedBrick AI servers. Direct Upload is commonly used for light experimentation with a small dataset.
All image data directly uploaded will be stored on RedBrick AI servers. If you don't want RedBrick AI to host your image data, please integrate your storage.
RedBrick AI supports a variety of different image formats:
  1. 1.
    DICOM - .dcm, .ima, .dicom, .dicm
  2. 2.
    NIfTI - .nii, .nii.gz
  3. 3.
    Videos - .mp4, .mov, .avi
  4. 4.
    RGB Images - .png, .jpeg, .jpg, .bmp
Navigate to the project you want to upload data within to upload your scans. Click on "Upload Data" on the top-right of the dashboard.
  1. 1.
    Select the type of data you want to upload. Note - you can only upload one type of image data (DICOM, NIfTI, etc.) at one time, and each data type has its folder structure format.
  2. 2.
    For DICOM & NIfTI volume data, select yes/no for group by study. Grouping by study allows you to upload multiple scans as a single task. This is useful when you want to view/annotate multiple images (a full study) at once.

Folder Structure

DICOM 3D Volume

Put all instances of a DICOM series in a folder. If you are uploading only 1 series, you can upload all the instances, not in a folder.

NIfTI 3D Volume

Individual NIfTI files are uploaded as separate tasks. If you want to group tasks by study, put the NIfTI files of a single task/study in a folder.

Image 2D

Individual 2D images are all uploaded as individual tasks. If you want to create a study task with 2D images, please use your external storage, or upload data using the CLI.

Video Files

Individual 2D videos are all uploaded as individual tasks. If you want to create a study task with 2D videos, please use your external storage, or upload data using the CLI.

Video Frames

Group all frames for a video in a single folder. The video frames will be naturally sorted using the frame filenames.