[Loris-dev] importing MINC files into LORIS - registerFile.pl script ?

Wed Jun 28 11:54:44 EDT 2017

Hi Sridar,

Yes, I can confirm the script already exists. It is called 'batch_uploads_imageuploader'. It operates on a zip of DICOMs (not on MINCs) without using the imaging uploader GUI. It takes in a standard in (text file) of the entries (i.e. a zip of DICOM scans for a given candidate's MRI session) that need to be inserted into the database.

The scans should be placed on the filesystem and the text file would point to them, one at a time, and the above script will read the entries in the file, one line at a time, and process them for you from the backend without using the GUI.

Details of how the script is launched and what additional inputs are required can be found here:

https://github.com/aces/Loris-MRI/pull/133

We have a meeting today (which I believe is to discuss this).

- Mouna

________________________________
From: Sridar Narayanan, Dr.
Sent: Wednesday, June 28, 2017 11:43:11 AM
To: Mouna Safi-Harab
Cc: Istvan Akos Imre Morocz, Dr; Mouna Safi-Harab; loris-dev; Jordan Stirling
Subject: Re: [Loris-dev] importing MINC files into LORIS - registerFile.pl script ?

Hi Mouna,

While the import/upload of DICOMs via the API will be a welcome addition to the LORIS tools, I think the immediate need is for batch upload of MRI files. Samir has indicated that this ability is already available (batch_upload_mri). Could you confirm whether this script operates only on MINC files, or DICOM as well? I think Pisti will need to know the details of how to run this script, as we will be importing tens of thousands of already-acquired MRI data into LORIS for the IPMSA project.

Thanks,

Sridar

On Jun 28, 2017, at 7:24 AM, Mouna Safi-Harab <mouna.safi-harb at mail.mcgill.ca<mailto:mouna.safi-harb at mail.mcgill.ca>> wrote:

Hi Pisti,

When you register a MINC file into the database, you should be able to store whatever MINC header is in the MINC file itself. So if the conversion from DICOM to MINC resulted in some loss, this will be reflected in the database (in other words, the database can only store the extent of what is in the MINC file, and comparing this to a DICOM starting point is dependent on the dcm2mnc conversion, not Loris).

Regarding your second question, I know there is a few important things on the API development/improvement roadmap, one of which is uploading DICOMs and launching the pipeline through the API, however, I am not sure that end of summer is the projected completion time. I was under the impression that downloading DICOMs is more feasible by the end of the summer than uploading and inserting through the API, but I could be wrong. Xavier, Dave, and Samir which are on this email list can comment on the API timelines and priorities better than me.

- Mouna

________________________________
From: loris-dev-bounces at bic.mni.mcgill.ca<mailto:loris-dev-bounces at bic.mni.mcgill.ca> <loris-dev-bounces at bic.mni.mcgill.ca<mailto:loris-dev-bounces at bic.mni.mcgill.ca>> on behalf of István Ákos Mórocz <istvan.morocz at mcgill.ca<mailto:istvan.morocz at mcgill.ca>>
Sent: Tuesday, June 27, 2017 1:58:57 PM
To: Mouna Safi-Harab
Cc: loris-dev; Jordan Stirling; Sridar Narayanan, Dr.
Subject: Re: [Loris-dev] importing MINC files into LORIS - registerFile.pl script ?

dear Mouna,

one of the key questions for us is if we loose essential information
by importing MINC files into LORIS using the registerFile.pl script -
as compared to using DCM files via the regular=manual LORIS way.  i
assume here that we do a 'perfect' job outside LORIS when we convert
DCM files to MINC files by using our own scripts.

then, we just heard this morning that the next LORIS version this
summer will feature a DCM importing API - is that truly the case ?

bye, pisti

On Thu, 22 Jun 2017, István Ákos Mórocz wrote:

> sounds good, dear Mouna - we will
> give a try this coming tuesday in
> the afternoon !  thanks for your
> email !  bye, pisti
>
> On Thu, 22 Jun 2017, Mouna Safi-Harab wrote:
>
>> Hi Pisti,
>> I will send a more detailed answer after the long weekend, but for now, the
>> easiest way for you to know how the script functions is
>> by simply typing at the command line (from where the MRI codebase
>> directory):
>> registerFile.pl -help
>>
>> This applies to almost all .pl scripts we have in the codebase.
>>
>> From that, you can see all the arguments and options you have to be able to
>> run the script. But almost inevitably, you will have to
>> write a wrapper script that is specific to your project and how you want to
>> import the data (with some variable initialisation) that
>> itself will call the registerFile.pl script.
>>
>> Give it a try, and we can talk more after the weekend.
>>
>> - Mouna
>>
>> On Tue, Jun 20, 2017 at 12:47 PM, István Ákos Mórocz
>> <istvan.morocz at mcgill.ca<mailto:istvan.morocz at mcgill.ca>> wrote:
>>       dear LORIS developers,
>>
>>       as mentioned before in today's LORIS meeting, our group needs to
>> learn
>>       how to import some 42000 old MRI studies in MINC format into LORIS.
>>
>>       we learned in the mean time how to create CANDIDATE and VISIT entries
>>       in LORIS via its jason API.  because there is no API for DICOM upload
>>       (and because we may use our own dcm2minc pipeline) we were delighted
>>       to hear from Mouna last week that one may go another route and import
>>       MINC files actually directly into LORIS - so that subsequently Louis
>>       Collins et al will be able to deal with QA and other processing
>> steps.
>>
>>       question : can you refer us to weblinks, HOWTOs and any material on
>>       using your registerFile.pl script and related environment so we get
>>       this MINC upload mechanism working as soon as possible ?
>>
>>       thanks much, bye, pisti
>>
>>
>>         ,-*~'`^`'~*-,._.,-*~'`^`'~*-,._
>>        '
>>       :        http://noisis.com<http://noisis.com/>
>>        `    istvan.morocz at noisis.com<mailto:istvan.morocz at noisis.com>
>>        :        +1-514-927-1169
>>        :    istvan.morocz at mcgill.ca<mailto:istvan.morocz at mcgill.ca>
>>        :  http://cafe.spl.harvard.edu<http://cafe.spl.harvard.edu/>
>>        ,    http://www.pistikem.org<http://www.pistikem.org/>
>>       :       pistikem at gmail.com<mailto:pistikem at gmail.com>
>>        .
>>         `~*-,._.,-*~'`^`'~*-,._.,-*~'`"
>>
>>
>>
>>
>
>  ,-*~'`^`'~*-,._.,-*~'`^`'~*-,._
> '
> :        http://noisis.com<http://noisis.com/>
> `    istvan.morocz at noisis.com<mailto:istvan.morocz at noisis.com>
> :        +1-514-927-1169
> :    istvan.morocz at mcgill.ca<mailto:istvan.morocz at mcgill.ca>
> :  http://cafe.spl.harvard.edu<http://cafe.spl.harvard.edu/>
> ,    http://www.pistikem.org<http://www.pistikem.org/>
> :       pistikem at gmail.com<mailto:pistikem at gmail.com>
> .
>  `~*-,._.,-*~'`^`'~*-,._.,-*~'`"

   ,-*~'`^`'~*-,._.,-*~'`^`'~*-,._
  '
:        http://noisis.com<http://noisis.com/>
  `    istvan.morocz at noisis.com<mailto:istvan.morocz at noisis.com>
  :        +1-514-927-1169
  :    istvan.morocz at mcgill.ca<mailto:istvan.morocz at mcgill.ca>
  :  http://cafe.spl.harvard.edu<http://cafe.spl.harvard.edu/>
  ,    http://www.pistikem.org<http://www.pistikem.org/>
:       pistikem at gmail.com<mailto:pistikem at gmail.com>
  .
   `~*-,._.,-*~'`^`'~*-,._.,-*~'`"

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.bic.mni.mcgill.ca/pipermail/loris-dev/attachments/20170628/ea9888c5/attachment.html>