[Vv] Fwd: Bug in clitkDicom2Image

Thu Dec 20 13:34:30 CET 2018

Hi Thomas,

I have a segmentation fault error (Segmentation fault: 11) with both the 
-p option and without.

Ben

Le 20/12/2018 à 13:17, Thomas BAUDIER a écrit :
> Hi Benoit,
>
> I did a modification:
> https://github.com/open-vv/vv/commit/133a3b14fe0d225aff4ea4edd53d61e7f3d063c2 
>
>
> In fact, the processes to open dicom series with vv and with 
> clitkDicom2Image are different. Now, clitkDicom2Image does the same 
> thing than vv (using itk::GDCMSeriesFileNames).
>
> Does it work for you ?
>
> Thomas
>
> On 2018-12-19 17:58, Benoît Presles wrote:
>> Dear Thomas,
>>
>> I have just checked the new version and there is still an error either
>> in vv or in clitkDicom2Image (I think in clitkDicom2Image).
>> First, I convert the dicom image I sent you WITHOUT the -p option and
>> open it in vv. Then, I open in vv (vv-->Open Dicom) the dicom image
>> without activating the "Patient Coordinate System" checkbox and I link
>> the two images. They are NOT aligned.
>>
>>
>> Thanks for your help,
>> Best regards,
>> Ben
>>
>>
>>
>> Le 19/12/2018 à 16:09, Thomas BAUDIER a écrit :
>>>
>>> Hi Benoit,
>>>
>>> I think it's ok now with this commit:
>>> https://github.com/open-vv/vv/commit/5f801bf0b07486889123e941d6913d4369dfc86f 
>>> Thank you
>>> Thomas
>>>
>>> On 2018-12-18 08:47, Thomas BAUDIER wrote:
>>>> Hi Benoit,
>>>>
>>>> Ok, I will investigate the problem. Yes, could you send me a dicom
>>>> image please. You can send it directly to my email if you do not want
>>>> to send it to the mailing list.
>>>>
>>>> Thank you
>>>> Thomas
>>>>
>>>> On 2018-12-17 20:00, Benoît Presles wrote:
>>>>> Dear Thomas,
>>>>>
>>>>> I updated to the latest version of vv.
>>>>>
>>>>> About vv, my bad. I did not notice the "Patient Coordinate System"
>>>>> checkbox in the new version. When I open a dicom with this checkbox
>>>>> activated, I get the same image orientation as with the old version.
>>>>>
>>>>> About clitkDicom2Image, I confirm you that the -p option does not
>>>>> work. When I use it, the image I generate has no transformation 
>>>>> matrix
>>>>> and it is neither aligned with the image I generated with the old
>>>>> version of clitkDicom2Image.
>>>>> Moreover, when I don't use the -p option, the transformation 
>>>>> matrix is
>>>>> the identity but the origin of the generated image is not the same as
>>>>> the dicom image, so it is not aligned with the dicom image open
>>>>> without activating the "Patient Coordinate System" checkbox.
>>>>>
>>>>>
>>>>> Do you want me to send you an image in dicom format with a 
>>>>> transformation ?
>>>>>
>>>>>
>>>>> Best regards,
>>>>> Ben
>>>>>
>>>>>
>>>>>
>>>>> Le 17/12/2018 à 17:10, Thomas BAUDIER a écrit :
>>>>>>
>>>>>> Hi Benoit,
>>>>>>
>>>>>> Sorry for the late reply, I have a lot of work :/
>>>>>>
>>>>>> I do not know your previous version of clitk but I did some 
>>>>>> modification to take into account ITK transformation management 
>>>>>> modification.
>>>>>> You can see the commits here:
>>>>>> https://github.com/open-vv/vv/commits/master/tools/clitkDicom2Image.cxx 
>>>>>> But I do not know why with the -p option it does not work anymore.
>>>>>>
>>>>>> But when you open the dicom serie with vv, you can set "Patient 
>>>>>> Coordinate System" to take the transformation into account when 
>>>>>> you select the serie. Does it work for you ?
>>>>>>
>>>>>> Best regards
>>>>>> Thomas
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> On 2018-12-17 12:34, Benoît Presles wrote:
>>>>>>> Dear Thomas,
>>>>>>>
>>>>>>> Did you have time to have a look at the bug ?
>>>>>>>
>>>>>>> Best regards,
>>>>>>> Ben
>>>>>>>
>>>>>>>
>>>>>>> Le 11/12/2018 à 15:25, Benoît Presles a écrit :
>>>>>>>> Dear Thomas,
>>>>>>>>
>>>>>>>> I have updated my version of clitk and a bug I had before 
>>>>>>>> reappeared!
>>>>>>>> clitkDicom2Image does not take into account the transformation 
>>>>>>>> in the dicom header (Image Orientation (Patient) tag in the 
>>>>>>>> dicom header) even with the -p option (open the image with 
>>>>>>>> patient coordinate system).
>>>>>>>> Similarly, when I open the dicom image with vv (vv-->Open 
>>>>>>>> Dicom), the transformation is not taken into account.
>>>>>>>>
>>>>>>>> In a previous version of vv, the transformation was correctly 
>>>>>>>> taken into account.
>>>>>>>>
>>>>>>>>
>>>>>>>> Thanks for your help,
>>>>>>>> Best regards,
>>>>>>>> Ben
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Le 24/07/2017 à 16:50, Thomas BAUDIER a écrit :
>>>>>>>>> Hi Benoit,
>>>>>>>>>
>>>>>>>>> I just came back from hollidays and I see your bug.
>>>>>>>>> Thank you to your support, I checked your files and I 
>>>>>>>>> validated your modifications.
>>>>>>>>> So they are available on the git.
>>>>>>>>>
>>>>>>>>> Best
>>>>>>>>> Thomas
>>>>>>>>>
>>>>>>>>> On 2017-07-20 20:21, Benoît Presles wrote:
>>>>>>>>>> Dear vv users,
>>>>>>>>>>
>>>>>>>>>> Please find attached the fix I did.
>>>>>>>>>>
>>>>>>>>>> Best regards,
>>>>>>>>>> Ben
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> -------- Message transféré --------
>>>>>>>>>> Sujet :     Bug in clitkDicom2Image
>>>>>>>>>> Date :     Wed, 12 Jul 2017 09:39:24 +0200
>>>>>>>>>> De :     Benoît Presles <benoit.presles at u-bourgogne.fr>
>>>>>>>>>> Pour :     vv at creatis.insa-lyon.fr
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> Dear Thomas,
>>>>>>>>>>
>>>>>>>>>> I think there is a bug in the clitkDicom2Image tool. 
>>>>>>>>>> clitkDicom2Image
>>>>>>>>>> does not take into account the transformation in the dicom 
>>>>>>>>>> header
>>>>>>>>>> (Image Orientation (Patient) tag in the dicom header).
>>>>>>>>>> Indeed, the same two images, the first one in "dicom format"
>>>>>>>>>> (vv-->Open Dicom) and the other in "image format" generated 
>>>>>>>>>> thanks to
>>>>>>>>>> clitkDicom2Image, do not overlap.
>>>>>>>>>>
>>>>>>>>>> You can download my test data here:
>>>>>>>>>> https://cloud.u-bourgogne.fr/index.php/s/dNkQJuvXepdmx7e. You 
>>>>>>>>>> will
>>>>>>>>>> find a dicom image with an orientation and the corresponding 
>>>>>>>>>> image
>>>>>>>>>> generated thanks to clitkDicom2Image.
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> Thanks for your help,
>>>>>>>>>> Best regards,
>>>>>>>>>> Ben
>>>>>>>>> _______________________________________________
>>>>>>>>> vv mailing list
>>>>>>>>> vv at creatis.insa-lyon.fr
>>>>>>>>> http://www.creatis.insa-lyon.fr/mailman/listinfo/vv
>>>>>>>>
>>>>>>>>
>>>>>>>> _______________________________________________
>>>>>>>> vv mailing list
>>>>>>>> vv at creatis.insa-lyon.fr
>>>>>>>> https://www.creatis.insa-lyon.fr/mailman/listinfo/vv
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> vv mailing list
>>>>>>> vv at creatis.insa-lyon.fr
>>>>>>> https://www.creatis.insa-lyon.fr/mailman/listinfo/vv
>>>>>> _______________________________________________
>>>>>> vv mailing list
>>>>>> vv at creatis.insa-lyon.fr
>>>>>> https://www.creatis.insa-lyon.fr/mailman/listinfo/vv
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> vv mailing list
>>>>> vv at creatis.insa-lyon.fr
>>>>> https://www.creatis.insa-lyon.fr/mailman/listinfo/vv
>>> _______________________________________________
>>> vv mailing list
>>> vv at creatis.insa-lyon.fr
>>> https://www.creatis.insa-lyon.fr/mailman/listinfo/vv
>>
>>
>> _______________________________________________
>> vv mailing list
>> vv at creatis.insa-lyon.fr
>> https://www.creatis.insa-lyon.fr/mailman/listinfo/vv
> _______________________________________________
> vv mailing list
> vv at creatis.insa-lyon.fr
> https://www.creatis.insa-lyon.fr/mailman/listinfo/vv