Strange MD and FA values in preclinical rat and mice data in DSI studio

[Javier Arcos Hódar]

Greetings DSI studio team

I'm Javier Arcos Hódar, a research assistant based on the Alberto Sols Institute for Biomedical Research on Madrid. I've been tasked to investigate other software for MRI data processing in case we end up having problems with our homemade Matlab software. First of all I'd like to be thankful for the developing of DSI Studio, since it seems to work faster and provide many options ours doesn't. Unfortunately I might be mismanaging something since even for the very simple task I wanted to start with (exporting AD, MD and FA values) it's giving me some problems. I attached to the dropbox the relevant files, from both DSIstudio (src, fib and exported maps) the raw Bruker data and the results from our software.

The files from our alternative software are not a .nii stack but can be opened on imageJ under Import -> Raw -> DTI_FA_Raw, with settings (I think those are the default but please double check to avoid problems) -image type 64-bit Real -Width and Height 128 pixels, -offset of first image 0, -number of images 1, -gap between images 0 bytes, and -only little-endian byte order ticked on, with the three other settings on blank.

I think it is evident (both from checking with my group and the literature, see for example Li XY et al (2017) https://www.sciencedirect.com/science/article/pii/S1995764517304492) that the high FA values in the .nii file DSI Studio creates cannot be correct, with the typical in slice 1 close to 0.60 in some areas (0.17 for contralateral region with no tumour and 0.31 for tumour are the values cited in the literature, our own are similarly in that range, mostly in the .20s for non-tumour areas) but it might be just that I'm misusing DSI Studio in some way rather than a problem with the software itself. My settings for reconstruction are T2b(1) DTI, T2b(2) check b-table and tensor matrix marked on while helix angle and scheme balance are marked off. While the b table and directions are correctly read from the 2seq file of the raw Bruker data it should be noted that when setting up the mask in the src file I also edit ->image flip X and image flip Y to fix the orientation in the final .nii export, but this shouldn't matter when it's just applying the FA formula pixel per pixel. 

Initially I thought the MD values could be quite closer to reality but on the wrong unit (multiplying the values associated with each voxel by 1000 in R gave me values with a near-unity correlation with the processing of our other sofware, if always somewhat lower, our software gives MD values for each voxel in μm^2/s) and that could even be the source of the error (as in FA values would be correct if I could somehow change DSI studio's processing to also multiply all MD valuse by 1000) but now I'm beginning to doubt even that.

While attached is an example I could provide data from other subjects if necessary to narrow down the issue.

Thanks in advance

[Frank Yeh]

I will check and get back to you.
Frank

> --
> You received this message because you are subscribed to the Google Groups "DSI Studio" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to dsi-studio+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/a277b712-9453-49cf-b655-01ceea5bfbf7o%40googlegroups.com.

[Javier Arcos Hódar]

Thanks for the quick warning. Once thing I forgot to mention that I’d also like to point out we have two basal images per diffusion weighted slice (as would be evident converting the 2seq to dicom or nii), which I don’t rule out could lead to the issues if not recognized by DSI Studio.

 

Javier

 

Enviado desde Correo para Windows 10

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/CAG_QrtJ1TeGWeWfi%2Br9WNDfbDsEG%2BXszsTDxagJN1u29HMTMDg%40mail.gmail.com.

 

[Frank Yeh]

The FA values in some slices seem correct with FA values mostly lower
than 0.6, but in others the reconstruction results are bad.
This indicates a high possibility of acquisition issues.
The raw DWI in [Step T2] shows a lot of motion artifacts. It is likely
that the high FA is due to this problem because the acquisition has
only 7 directions in each of the b=300 and 1000 acquisition.

I did not find any problems other than the motion artifacts.

I would suggest you try other DTI software and see if they can better
handle signal dropout issue due to motion.

Best regards,
Frank