How to structure the data after processing

[etetteh]

I have downloaded and processed all the datasets following the instructions from the docs.
However the structure for example the data_processed/sub-amu01/anat/ looks like the image below.

I believe I don't need all the files and would like to know the exact files needed to perform a segmentation task using ivadomed. Thank you, and counting on your usual support.

[jcohenadad]

Hi @etetteh, thank you for reaching out. The analysis you ran does perform some segmentations (e.g., of the spinal cord on multiple contrasts, and of the spinal cord gray matter on the T2*w contrast), however it is not relying on ivadomed. What exactly do you have in mind? Would you like to train a model with ivadomed to perform a segmentation of the spinal cord? If this is the case, then, depending on the scope of your project and architecture (e.g., single vs. multiple contrasts, already co-registered or not, what input resolution, etc.), you might need to adjust the analysis script. Also, you need to consider what kind of ground truth segmentation you would like to work with. So as you see there is a lot of clarifications needed to be able to properly address your question.