to3d usage question

AFNI Message Board
1

Hi, List Experts,

Could you advise how I can use to3d to convert DICOM to nii format? I attempted to use the command.

to3d -prefix EPI_Sub1 -time: zt nz nt TR alt+z Input_DIC

I am NOT clear how to know whether my slices are in the order “z-axis first, then t-axis” or the contrary “t-axis first, then z-axis”. Also, how to decide whether I should use alt+z, or alt+z2…? I read the help but am still unclear. Could I directly write the code adding two new arguments?

to3d -prefix EPI_Sub1 -time: zt nz nt TR alt+z - assume_dicom_mosaic -save_outliers outlier_fname Input_DIC

Thanks a lot,
Juan

2

For my data, I find some information

Private_0051_100e: ‘Sag>Tra(2.1)’ for one data set
Private_0051_100e: ‘Sag> Cor (-2.7)>Tra(-0.0)’ for another dataset.

I guess that they are some relevant to the z-axis or t-axis, but not confident. Any comments are much appreciated.

Juan

3

I am still checking the relevant information myself here.

4

Hi Juan,

To be clear, what is Input_DIC, a wildcard patter for a
list of files?

Consider something like:

Dimon -dicom_org -gert_create_dataset -gert_write_as_nifti -infile_pattern Input_DIC
  • rick
5

Hi, Rick,

Thanks so much for your prompt feedback.

Sorry, those are the dcm files (DICOM). They can be either functional or anatomical dcm files. for instance, in my folder" anatomical", there are many dcm files: image00005.dcm,image00006.dcm,image00007.dcm,image00008.dcm,image00009.dcm,image00010.dcm,…image00150.dcm,

I just tried using the command below,and got outputs: dimon.files.run.010; GERT_Reco_dicom_010;OutBrick_run_010.nii

Dimon -dicom_org -gert_create_dataset -gert_write_as_nifti -infile_pattern ‘anatomical/*.dcm’

Several questions: What is the file “GERT_Reco_dicom_002” for? I got some warnings:

*+ WARNING: Slice-based center is different from mosaic center
*+ WARNING: Origin computation of obliquity may be incorrect
Mosaic Center : 2.0819 -9.08766 45.4386
Slice-based Center: 3.58128 -7.88988 44.5345

Significant outliers detected in these sub-bricks:
0 1 2 4 45 46 52

How should I handle this, or just ignore this? I have not tried “3dToutcount -save outnam dataset | 1dplot -stdin” in the past. Any good pointers?

I am noted that “if images are wrong, try using ‘-assume_dicom_mosaic’”, how should I use this?

Thanks so much,
J

6

Hi, Rick,

I have searched some old posts from you and other list users. I get to know that Dimon would extract the relevant information from the DICOM file and provide it to the to3d for the process. Given some issues, I tried to include two options (i.e., - assume_dicom_mosaic, -skip_outliers) to the Dimon command, but it does not work. To get clear what the exact process is, could you offer me some instruction on those issues? For instance, if we use the command below.

to3d -prefix EPI_Sub1 -time: zt nz nt TR alt+z - assume_dicom_mosaic -skip_outliers -save_outliers outlier_fname Input_document_dicom

How to know whether my slices are in the order “z-axis first, then t-axis” or the contrary “t-axis first, then z-axis”. Also, how to decide whether I should use alt+z, or alt+z2 (I know the difference between them)? I ask you these questions as I myself can not find the information from the dicom file headers. I am also not clear how Dimon could get them and then provide it to to3d function - I tracked the output from the Dimon process, but could not find it. It seems that no one has made it clear.

Thank you very much for your time,
J

7

It is hard for me to say what is needed here.
That is really for someone that knows the
scanner sequence to verify.

But please add the Dimon option,
-save_details DET

and mail me the DET* text files.

Thanks,

  • rick