NITRC Automatic Registration Toolbox Forum: help

using ART2.0 PIL transform

Matthieu Faillot — Wed, 19 Apr 2023 12:34:07 GMT

HI
I am trying to use PIL transformation in FSL format to reorient masks in ACPC RAS space.
Applying the PIL transform to the native T1 image (acpcdetect input) doesn't give the same result as the acpcdetect output (t1_RAS.nii).

I tried to use the ART reslice program to apply the mrx file to the native T1 but the result is different from acpcdetect output (and different from the output of flirt). It is also strange that when using relisce with the output mrx transform from yuki (with the same t1 image as input), the output is different that when using the transform from acpcdetect (image is completely flipped).
I also tried converting the initial T1 to PIL orientation (with fslswapdim) before applying PIL transformation but there is still a difference between the output of flirt and the output of acpcdetect (t1_RAS.nii)
Is there a simple method to apply the PIL transform from ACPCdetect to masks using flirt (or another program) ?
The only workaround I found so far is to comput a new transform with flirt between the native T1 image and the output of acpcdetect (t1_RAS.nii)
Best regards
Matthieu

CC subregions MNI space output?

Wesley Thomas — Tue, 14 Feb 2023 23:30:31 GMT

Hi,

When I segment the CC, I get a 2D 512x512 nifti, is there a way for it to output a 3D atlas in MNI space?

Thanks,
Wesley

Please make art2 available so than I can perform functional to structural co-registration

Roger Tait — Thu, 07 Jul 2022 11:42:58 GMT

Hi All,

Please make art2 available so than I can perform functional to structural co-registration.

Roger Tait

RE: Working with the cc subregions

Matthieu Faillot — Wed, 15 Jun 2022 17:03:50 GMT

Dear Dr Ardekani
I am using yuki2.1 and it works great
I wonder if there are any updates which allow to output the transform from native space to midsagittal plane when using the software
This would help to visualize the callosal segmentation in native space.

Best regards
Matthieu

Cannot open mdl. file

Carina Forster — Thu, 07 Jan 2021 10:26:22 GMT

Hello community,

I encounter the following error using acpc detect:

acpcdetect -i $ARTHOME/sub_02/T1.nii -v

I cannot open the model file: /data/u_forster_software/art/T1acpc.mdl.

I used acpc detect before and it worked so I am not sure what happened.

Thank you for your help.

Carina

Edit:

It is somehow related to my remote working client, made it work again, thank you for this great tool!

atra cannot find orion.mdl

tom parker — Fri, 17 Apr 2020 14:06:57 GMT

Dear all,

I was running atra for AC-PC realignment using the following command: atra -i imagelist -v -o foo

Everything seemed to be going well up to the point in which the following warning came out: Could not find /usr/local/art/orion.mdl, aborting ...

I have checked the art folder and the file orion.mdl is indeed there. Does anyone have any ideas why this is happening?

Thanks a lot!

empty average file after atra

valeriarm — Wed, 15 Apr 2020 22:48:23 GMT

Hi!

Thanks to the float -> 'short' or 'unsigned character' from William Quinn, I was able to run atra. But, when looking to the average file it is empty.

In the metadata I can see that dimensions 4,5 and 6 are different to zero, and that concern me.

Please help!!!

Thanks in advance

Valeria :3

RE: cropped anatomical image after acpcdetect

Babak Ardekani — Tue, 17 Dec 2019 16:19:18 GMT

[i]Originally posted by sam1:[/i][quote]Hello,

I've used the command acpcdetect -i anat.nii on my anatomical image, and the outputted anat_RAS.nii came cropped at the top of the head. Is this to be expected? I tried playing with the different options but the image produced is always slightly cropped at the top...
When I load both the anat and anat_RAS in fsleyes both images have the same dim1, dim2 and dim3, but they have different X,Y, Z sizes.

Best regards,
Sam[/quote]
[color=#000000]Hi Sam,[/color]

[color=#000000]Is it possible that anat.nii did not cover the entire head to begin with? Another possibility is that MSP and AC/PC detection simply failed for this case. You can check that by looking at the *PPM images which show the detected landmarks and the MSP. Regarding dimensions, acpcdetect allows you to explicitly specify the output dimensions (both matrix sizes and voxel sizes). [/color]

[color=#000000]Babak[/color]

cropped anatomical image after acpcdetect

sam1 — Tue, 17 Dec 2019 15:53:39 GMT

Hello,

I've used the command acpcdetect -i anat.nii on my anatomical image, and the outputted anat_RAS.nii came cropped at the top of the head. Is this to be expected? I tried playing with the different options but the image produced is always slightly cropped at the top...
When I load both the anat and anat_RAS in fsleyes both images have the same dim1, dim2 and dim3, but they have different X,Y, Z sizes.

Best regards,
Sam

RE: Installation problem

Babak Ardekani — Mon, 19 Aug 2019 16:30:48 GMT

[i]Originally posted by lll55555:[/i][quote]Hi.

I found the 3dwarper, and downloaded it.
1. I moved 3dwarper file to /usr/local/art/bin/
2. changed the permission on this file as -rwxrwxr-x

When I typed 3dwarper -h, it says there is not such command.
But, it works, typing ./3dwarper -h in /usr/local/art/bin.

I'm new to linux. Can u help me out?

ps.
I set environment variable like this as the instruction, installing acpcdetect.

Set the $ARTHOME environment variable. To set $ARTHOME, csh and tcsh users should
add this line in their '.cshrc' or '.tcshrc' files:

setenv ARTHOME /usr/local/art
sh and bash users should instead add these lines in their '.profile' or '.bashrc':
ARTHOME=/usr/local/art
export ARTHOME[/quote][color=#000000]Seems like you have gotten it running properly on your system. Please read the documentation on 3dwarper which is also online.[/color]