Links

Troubleshooting

This section covers troubleshooting issues in loading data in the annotation tool. If there is an error in loading your data, you will encounter the following error screen.

Troubleshooting

  1. 1.
    You may have a specific error message under "Failed to load image." This error message could provide insight into why your task is not loading. In the image above, "Request failed with status code 404" likely means an external storage integration or internet connection issue.
  2. 2.
    If you're using external storage and can see URLs enumerated under "Failed to load these URLs," try opening the link in your browser.
    1. 1.
      If the browser loads a screen with an error message, there is likely an issue with your Storage Method integration (possibly your bucket name or access keys). You can double-check the validity of your storage method integration using the "Verify" function within the Storage Method tab. It is also possible that your Storage Method integration is valid, but the item's path you have provided is invalid, i.e., the data point doesn't exist in your bucket.
    2. 2.
      If the browser downloads the data file (or displays it, in the case of .png, .jpeg, etc.), you probably have not enabled CORS on your bucket. If CORS is not enabled on your bucket, your browser will not be able to load the image (because it originates from a different origin).
  3. 3.
    If you have double-checked the validity of your Storage Method integration & items paths and enabled CORS on your bucket, the data loading issue is likely because of the image file's contents. Here are some possible problems:
    1. 1.
      Your DICOM or NIfTI files may not be wholly defined, i.e., have all required headers for us to display the image.
    2. 2.
      If you see the error message "Set maximum size exceeded," this means you have uploaded an annotation file with an instance ID greater than 256. This could happen if you accidentally import an image file in the segmentations field in the items file.
    3. 3.
      Our interface is expecting NIfTI files of type UInt8. If you import image or annotation files with floating point values, you will get the following error message "Values must be integers."

What to do next

If you're unable to resolve the data loading issue, we recommend the following steps:
  1. 1.
    Click on the "Raise Issue" button. This will remove the task from your labeling queue, preventing it from appearing repeatedly. Learn more about raise issue here. Once you raise an issue, you will automatically be taken to another task in your queue.
  2. 2.
    If you don't want to raise an issue, navigate to another task in your queue. The other tasks in your queue will be at the bottom of the error screen.