DTI images upside-down

[Ricardo Nicolas]

Hi!

I'm a neurosurgeon. I've done a dozen DTI studies by using DSI Studio with the same protocol and the same machine (Siemens Prisma) and your app helps a lot! But this time the tracking comes out as an inverted image (upside-down). The patient has been scanned twice (first time on Wednesday, second time on Friday). And each time it comes out as the same. We haven't modified any parameters on the MRI machine (as the MRI technician said). Please help! The patient is soon to have surgery.

PS. The detailed files have been uploaded to the dropbox for your further information.

MUCH THANKS!

The file index is in the screenshot. Weird is that there's no folder named after "MIX" in these scans. We used to have FA files in the MIX folder, and we used MPRAGE folder to insert and combine. This time we have no MIX folder. Thus we used the P2_0006 folder.

Tracking info:

The diffusion images were acquired on a SIEMENS Prisma scanner using a 2D EPI diffusion sequence (ep2d_diff_mddw_20_p2). TE=95 ms, and TR=3700 ms. A DTI diffusion scheme was used, and a total of 60 diffusion sampling directions were acquired. The b-value was 1000 s/mm2. The in-plane resolution was 1.71875 mm. The slice thickness was 5.2 mm. The restricted diffusion was quantified using restricted diffusion imaging (Yeh et al., MRM, 77:603–612 (2017)). The diffusion data were reconstructed using generalized q-sampling imaging (Yeh et al., IEEE TMI, ;29(9):1626-35, 2010) with a diffusion sampling length ratio of 1.25. A deterministic fiber tracking algorithm (Yeh et al., PLoS ONE 8(11): e80713, 2013) was used. A seeding region was placed at whole brain. The anisotropy threshold was 0.223596. The angular threshold was randomly selected from 15 degrees to 90 degrees. The step size was randomly selected from 0.5 voxel to 1.5 voxels. Tracks with length shorter than 30 or longer than 300 mm were discarded. A total of 50000 seeds were placed. parameter_id=50F6643E9A99193Fb803FdbF041b964350C3ecb

[Fang-Cheng 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/a0a32d26-10f8-4cde-83d1-442685daa487%40googlegroups.com.

[Fang-Cheng Yeh]

The raw DWI dicom already shows up-side down order of the slices.

To fix it in DSI Studio, in STEP T2 reconstruction, use [Edit][Image Flip z]

On Mon, Mar 30, 2020 at 12:15 AM Ricardo Nicolas <fauxv...@gmail.com> wrote:
>

[be weglichkeit]

Hey,

We have the same problems. I recognized that past reconstructions before march 2021 worked pretty well. Than an update of the Siemens PRISMA came in spring 2021 and since then we are not able to registrate the DTI with anatomical sequences. We tried several hundrets of combinations like flipping and swapping DTI but even the recommended "flip z" did not work well. Interestingly we saw that "the in-plane resolution was 1.5...mm" and "thickness 5.2" and we realized that these to parameters were opposite before march 2021.

Is there any further idea how to registrate both sequences to each correctly manually??? It seems that even axial, sagittal and coronal pictures are swapt and flipped. Do you plan an update to fix this? We would be very thankful, because a lot of data is already waiting for analysis ;)

Thank you very much

Sebastian Vetter (sebastia...@uni-leipzig.de)

Biomechanics

Sports Sciences, University Leipzig, Germany

[Frank Yeh]

I would suggest resampling the data to isotropic using [Edit][Resample isotropic] to see if it helps.

If not, please feel free to send me a sample of your data, and I can help figure it out.

Frank

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/47cf9d79-d4de-4510-8109-157bf7151021n%40googlegroups.com.

[be weglichkeit]

Dear Frank,

Thanks Frank,  even isotropic doesnt work. I uploaded my files (folder: vetter). we are interested in tractography for the shoulder (f.e. musculus supraspinatus, attached as region of interest for DSI Studio)

Cheers,

Sebastian Vetter

[be weglichkeit]

Dear Frank,

did you get the files I uploaded with Dropbox? I would be very happy if you might be able to find any solution and if you will be able to respond me via my official mail adresse (sebastia...@uni-leipzig.de)

Cheers,

Sebastian

[Frank Yeh]

Hi Sebastian,

     I can generate tractography: You may update DSI Studio (dated today) and process the data again from DICOM to get a new SRC file.

   Then followings the steps:

(1) Click [Step T2][Reconstruction] open 20210608_M030Y_Vetter_Sebastian_s04_ep2d_diff_mddw_12_p2.src.gz
(2) Set [Step T2a][Threshold]=0
(3) Set [Step T2b][Check b-table]=0
(4) Set [Step T2b(1)]=GQI
(5) Set [Step T2b(1)][Diffusion sampling length ratio]=1.25
(6) Click [Step T2b][Run reconstruction]

Hope this works for you.

Frank

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/e3e2cdcf-ee30-4383-8c4e-acb4f927b47fn%40googlegroups.com.

[Richard Ingram]

It could be the order of FA raw images. Try to read the images from an opposite order from what you read right now. 

On Dec 2, 2021, at 22:56, Frank Yeh <fran...@gmail.com> wrote:



To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/CAG_Qrt%2B642%2B%2Bh%2BghG4t2717nvPjpBnCzpoXPbpyYDpp5MM1AGQ%40mail.gmail.com.

[be weglichkeit]

Thank you very much, you are very helpful! With the latest version of DSI studio, it does work properly and even better with "step T2b(1) = DTI" for this set of data. 

But now I have the problem, that it is still the same upside-down/flipping thing with the latest data from october. Do I have to wait some time for another update fixing this? How can I change order of FA raw images? 

Best,

Sebastian Vetter

[Frank Yeh]

[Step T2][Edit][Image Flip z]

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/308f4520-e90f-49da-8d47-c1eaf048cc85n%40googlegroups.com.