Dti and dwi differences

MRI Acquisition
1

Hi team,
I have two quick question. what is the difference between dwi imaging and dti? how can I use spinal cord toolbox to convert dwi to dti? I appreciate if you can give me some hints, so I can work with some of my dwi data using your toolbox?
and can we consider dti a 4d image and other mri images like t2, t1, etc. 3d? thanks.

2

Hi @Roberto_r,

Thank you for reaching out!

Diffusion-weighted imaging (DWI), refers to the acquisition setup (i.e.: MR imaging with diffusion-encoding gradients), while diffusion tensor imaging (DTI) refers to the mathematical model applied to the DWI data.

I recommend reading these two articles, which explain the difference between DWI and DTI:

After reading these articles you should be able to answer your other questions (“can we consider dti a 4d image”).

Cheers,
Julien

3

thank you :slight_smile: I will read the papers in a proper time. I need a way to reach dti using dwi, can I make it using spinal cord toolbox ? and at the end there will be a new image (dti) that we can go for more processing using that?

4

I need a way to reach dti using dwi, can I make it using spinal cord toolbox ?

yes, more specifically with sct_dmri_compute_dti, which is powered by the awesome dipy.

and at the end there will be a new image (dti) that we can go for more processing using that?

Yes, DTI metrics will be output as separate NIfTI files.

1 Like
5

Thank you :slight_smile:
I installed your toolbox and used your above command to make nifitis of dti from dwi image (and by dti here we mean those fa, md, images … right?). it seemed nice to preprocess the dwi’s using spinal cord toolbox. just as a last question, is this okay or common using dti nifti’s for learning purpose? I had experience working with x-rays but not dwi image. is it better to give the network fa, md, … dti’s or the primary dwi image? thank you for your great work.

6

Hi!

I installed your toolbox and used your above command to make nifitis of dti from dwi image (and by dti here we mean those fa, md, images … right?).

right

is this okay or common using dti nifti’s for learning purpose?

it depends what is the purpose of the learning experience-- e.g. if your plan is to learn about basic MRI processing (eg: atrophy measure, voxel based morphometry) then DTI is not the right contrast. If your goal is to learn about diffusion-weighted imaging, then DTI is a good start.

is it better to give the network fa, md, … dti’s or the primary dwi image?

What network are you referring to? I need more context.

thank you for your great work.

Thanks! :blush:

7

right, I understood … I had written a small cnn network to detect tumor for my previous research, and was a bit curious to see if I can do something experimental with it using the dwi image or not. I want to do just a simple experiment.