Dear Peter,
Thanks so much for doing the experiment. This is very helpful and
ten~eleven hours should be good for us for now.
Best, Jieren
2017年9月29日星期五，Neher, Peter <p.ne...@dkfz-heidelberg.de> 写道：
> Dear Jiren,
>
>
> i repeated the simulation here and it also took about 11 hours.
> Unfortunately I can't change that right away, but I will look into it.
>
>
> Peter
>
>
>
> ------------------------------
> *Von:* Xu Jieren <jierenx...@gmail.com
> <javascript:_e(%7B%7D,'cvml','jierenx...@gmail.com');>>
> *Gesendet:* Mittwoch, 27. September 2017 17:56
> *An:* Neher, Peter
> *Cc:* mitk-users@lists.sourceforge.net
> <javascript:_e(%7B%7D,'cvml','mitk-users@lists.sourceforge.net');>
> *Betreff:* Re: [mitk-users] Generating the
>
> Dear Peter,
>
> We are currently using a Ubuntu 16.04 LTS, with i7Core, 11.7GiB memory,
> and 600+GB disk. I re-run the program yesterday, and it still took over ten
> hours. If you could help retest the simulation at your local, that would be
> really appreciated.
>
> Best, Jieren
>
> 2017-09-27 2:07 GMT-04:00 Neher, Peter <p.ne...@dkfz-heidelberg.de
> <javascript:_e(%7B%7D,'cvml','p.ne...@dkfz-heidelberg.de');>>:
>
>> Yes, unfortunately the simulations might take a while, depending on your
>> machine. But I will start the simulation here to double check if something
>> is going wrong. May I ask the specifications of the machine you are running
>> Fiberfox on?
>>
>>
>> Peter
>>
>> ------------------------------
>> *Von:* Xu Jieren <jierenx...@gmail.com
>> <javascript:_e(%7B%7D,'cvml','jierenx...@gmail.com');>>
>> *Gesendet:* Dienstag, 26. September 2017 18:29
>>
>> *An:* Neher, Peter
>> *Cc:* mitk-users@lists.sourceforge.net
>> <javascript:_e(%7B%7D,'cvml','mitk-users@lists.sourceforge.net');>
>> *Betreff:* Re: [mitk-users] Generating the
>>
>> Dear Peter,
>>
>> Thanks very much for helping. Yes, I have the same file downloaded.
>> After switching to the latest version, I have the file generated in a
>> nifti format. But this step took over 10 hours on my Ubuntu.
>>
>> I also figured out the reason for the dimension mismatch. I used a
>> vtk reader to invert the vtk ground truth file Fiber.fib to Matlab format
>> and it didn't interpret the dimension (~180*200*180) by the given hider.
>>
>>
>> Best, Jieren
>>
>>
>>
>>
>>
>> 2017-09-25 11:54 GMT-04:00 Neher, Peter <p.ne...@dkfz-heidelberg.de
>> <javascript:_e(%7B%7D,'cvml','p.ne...@dkfz-heidelberg.de');>>:
>>
>>> Dear Jieren,
>>>
>>>
>>> then you should switch to 2017.07.
>>>
>>>
>>> I am not sure which files you used to simulate your image, but the files
>>> provided on the challenge homepage (
>>>
>>> http://www.tractometer.org/downloads/downloads/ismrm_challen
>>> ge_2015/FilesForSimulation_v1.zip) include a mask of size 90*180*90.
>>> Probably you used the wrong mask. Where did you obtain the 180*200*180
>>> mask?
>>>
>>>
>>> Best regards,
>>>
>>> Peter Neher
>>>
>>>
>>>
>>> ------------------------------
>>> *Von:* Xu Jieren <jierenx...@gmail.com
>>> <javascript:_e(%7B%7D,'cvml','jierenx...@gmail.com');>>
>>> *Gesendet:* Montag, 25. September 2017 15:29
>>> *An:* Neher, Peter
>>> *Cc:* mitk-users@lists.sourceforge.net
>>> <javascript:_e(%7B%7D,'cvml','mitk-users@lists.sourceforge.net');>
>>> *Betreff:* Re: [mitk-users] Generating the
>>>
>>> Dear Peter,
>>>
>>> Thanks a lot for the reply. It helps a lot. Indeed I have been running
>>> the simulation in the version 2014 trying to have a faster implementation.
>>>
>>> And I'll clarify the second mask question. I noticed that the dimension
>>> of the ground truth bundles of the ISMRM 2015 data is around 180*200*180,
>>> with an input binary mask of the same dimension. However, the simulated
>>> ground truth nifiti files, i.e. the output of Fiberfox simulation, is of
>>> dimension 90*180*90 (parameters set for the generation signal), and this
>>> nifiti file has no output mask (which should be of dimension 90*180*90).
>>> Since this generated dwi data has no mask, I could not put a prior mask in
>>> its following processing steps, e.g. the QBall image generating step or the
>>> peaks extraction step. So I'm wondering if Fiberfox can output a generated
>>> tissue mask of the output dimension so that those following processure are
>>> possible, or is there a way that I could registrate the original mask to
>>> the generated data?
>>>
>>> Thank you.
>>>
>>> Best, Jieren
>>>
>>>
>>>
>>>
>>>
>>> 2017年9月25日星期一，Neher, Peter <p.ne...@dkfz-heidelberg.de
>>> <javascript:_e(%7B%7D,'cvml','p.ne...@dkfz-heidelberg.de');>> 写道：
>>>
>>>> Dear Jieren,
>>>>
>>>> which version of MITK Diffusion are you using? The current version
>>>> 2017.07 automatically outputs nifti files:
>>>> http://mitk.org/wiki/DiffusionImaging#Downloads
>>>> With the old version you should be able to load the .dwi file in the
>>>> GUI and save it as .nii, but I highly recommend updating to 2017.07.
>>>>
>>>> As for your second question, I am not sure what you mean by "QBall
>>>> generating pipeline". Are your referring to the Fiberfox simulation? The
>>>> mask for the generated file is the same as the mask for the simulation.
>>>>
>>>> Best regards,
>>>> Peter
>>>>
>>>> ________________________________________
>>>> Von: jierenxu <jierenx...@gmail.com>
>>>> Gesendet: Montag, 25. September 2017 02:31
>>>> An: mitk-users@lists.sourceforge.net
>>>> Betreff: [mitk-users] Generating the
>>>>
>>>> Dear developers,
>>>>
>>>> I've been trying to regenerate the ISMRM 2015 challenge data by the
>>>> files
>>>> provided on the challenge website using Fiberfox. And I went into the
>>>> following questions,
>>>>
>>>> 1. The output DWI data format is .dwi, however, the given challenge
>>>> data is
>>>> several .nii files and .bvecs and .bvals. Could I use MITK transform
>>>> between
>>>> those two data types? If not, how could I retrieve the challenge data in
>>>> .nii type?
>>>>
>>>> 2. There is a mask provided in the generating files. Could I have a
>>>> mask.nii
>>>> file of the generated .dwi data as well? Cause I notice that there are
>>>> options for a tissue mask for the QBall generating pipeline.
>>>>
>>>> Thank you!
>>>>
>>>> Best regards,
>>>>
>>>> Jieren Xu
>>>> Duke University
>>>>
>>>>
>>>>
>>>> --
>>>> Sent from: http://mitk-users.1123740.n5.nabble.com/
>>>>
>>>> ------------------------------------------------------------
>>>> ------------------
>>>> Check out the vibrant tech community on one of the world's most
>>>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>>>> _______________________________________________
>>>> mitk-users mailing list
>>>> mitk-users@lists.sourceforge.net
>>>> https://lists.sourceforge.net/lists/listinfo/mitk-users
>>>>
>>>
>>
>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
mitk-users mailing list
mitk-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mitk-users
