-
Notifications
You must be signed in to change notification settings - Fork 5
Frequently Asked Questions (FAQ)
David Bouget edited this page Jul 17, 2025
·
6 revisions
- As an initial test to verify that Raidionics works properly on your computer, and before trying using your own data, the following should be performed.
- Download the provided test data, available through the Raidionics GUI under
File > Download test data, and extract the content. - Load the two MR scans (one contrast-enhanced T1-weighted and one FLAIR) inside Raidionics and run the preoperative segmentation task, selecting "Contrast-enhancing" as tumor type. A tutorial video for this use-case can viewed on Youtube.
- Inspect the obtained segmentation masks if the processing succeeded. Otherwise please refer to the below list or visit the issues to find potential solutions.
- The information available in the Logs window might provide an explanation to an unsuccessful segmentation attempt, which can be found in the top menu-bar, under Settings > Logs (alternatively using the
CTRL + Lshortcut). - A purge of the models can be attempted to trigger a re-download. Simply open the Settings panel (
CTRL + Pshortcut), in System settings click the Models purge option (with a trash bin icon). ⚠️ By default, the acquisition sequence for each MR scan is automatically identified using a classification model. In case of incorrect classification, make sure the proper MRI sequence type is filled in correctly for each MRI series in the right-hand panel. In addition, the setting for using manually-defined sequences should be enabled (top menu-bar, under Settings > Preferences > Inputs/Outputs > Use manual MRI sequences).
- ❌ If the models are not being automatically downloaded when running Raidionics (e.g., internet/firewall issues), the segmentation and reporting tasks will not succeed.
- 💡 All the models used by Raidionics can be manually downloaded at the following Raidionics-models repo. After downloading each model archive, you simply need to extract the content and place the obtained folder inside
~/.raidionics/resources/models/. - ❗ Most tasks do require running multiple models, some as pre-processing steps. You can have a deeper look in the code where the executed pipelines are created, to download by hand the necessary models marked as download_model(model_name='MRI_Name')
- If you are running Raidionics on macOS, make sure you have installed the proper version of the software. The installer with a name trailing macOS.pkg will run on mac computers with Intel processors, while the installer with a name trailing macOS-14-15_arm64.pkg is meant to be used on mac computers with an ARM processor (M1/M2/M3 series).
⚠️ There is a specific installer (name trailing macOS-13_arm64.pkg) for the first generation of macOS with M1 chip (running macos-13).