Dcm2niix Troubleshooting
Introduction
The dcm2niix gear converts images in the DICOM format to NIFTI, and is often configured to run automatically as a Gear Rule.
Instruction Steps
This article gives some troubleshooting steps to take if you do not see your nifti file.
Verify the Gear Ran
To begin, see if the gear ran for your project:
- Sign in to Flywheel.
- Click Projects, and select the project where your data was added.
- Click Sessions.
-
Select the session you are interested in viewing.
-
Click the Provenance tab. If there is:
- No job for the acquisition: This means that the gear was not triggered to run. Follow the steps below to check the classifications. If you have this gear configured to run as a gear rule, you should also verify that the gear rule is configured correctly for your data.
-
A job with an error: Select View Log next to the job.
- If there is an error about being unable to find DICOM images, you should follow the steps below to verify the input file was accurately classified as a DICOM.
- If there are network errors, click Retry to rerun the gear.
Contact Flywheel support (Support@flywheel.io) if you have any questions about the log error messages
Verify Classifications
Classifications are used by Flywheel to consistently identify and label data. Flywheel uses these classifications (along with any other rules you have configured) to identify if files are eligible for the dcm2niix gear. If your files are not properly classified as DICOMs, then Flywheel will skip over them because it appears they are not valid input for this gear. In most cases, projects automatically classify new data by configuring the Scitran: DICOM MR Classifier gear as part of a Gear Rule. This gear extracts header information from the DICOMs to determine how to classify the image data. Learn more about best practices for data classifications in Flywheel.
Even if you have the DICOM MR Classifier gear set to run on new data, the classification process begins at the scanner. Your exam card should have all fields completed that are required for Flywheel. Learn more about how to name acquisitions for automated classifications. Note that the exact fields you need to fill out depend on your institution’s configuration. You should contact your local Site Admin for more details. To learn more, see our article about capturing data directly from MR scanners.
To verify your images have been classified:
- Sign in to Flywheel.
- Go to your project.
- Click Sessions.
- Select an session.
-
Select the Acquisitions tab.
-
Hover over an acquisition, and select the information icon .
-
Review the information. The following fields should all be configured:
- Modality
- Feature
- Intent
- Measurement
Here's an example of a classified file:
-
From here you can manually enter this information, and re-run the dcm2niix gear.
Verify Gear Rules
Check to make sure your gear rules are accurate for the data you are adding to Flywheel. To check the gear rules for your project:
- In your project, select the Gear Rules tab.
- Expand the dcm2niix gear rule.
- Verify the rules are correct for your data.
For example, check the version of the gear. If the gear automatically updated to the latest version, data evaluation may have changed. - Review the other gear rules for your project to make sure they don't conflict.
Help
Contact Flywheel support: Support@flywheel.io