Multi time point patient data.

83 views
Skip to first unread message

gaganjyoti sharma

unread,
Jun 25, 2015, 11:34:51 PM6/25/15
to xnat_di...@googlegroups.com
Hi all,

Please correct me if I am wrong. 

In our lab we have lot of longitudinal data, multiple time points (MR/CT) for the same patient.
I tried to push that data set towards XNAT. Now it was well received in pre archive, but archiving gave me conflict and when I tried to merge manually, the output was not good/favourable.
Ideally we want patient data archived in ID/Date_time/series format. Is it possible or I need to write a special pipeline for that ?

Regards
Gagan Sharma

PS: Any good step by step tutorial for pipeline, please post. 

Daniel Marcus

unread,
Jun 26, 2015, 12:06:07 AM6/26/15
to xnat_di...@googlegroups.com, gagan...@gmail.com
Hi Gagan,

XNAT is designed to support the sort of data representation you're looking for.  If you're getting a conflict on archiving, I suspect maybe you have the same value in the DICOM patient ID field for all of your sessions.  That's the field that XNAT uses to distinguish individual sessions.  Try putting different values in that field for the different DICOM studies and it will probably do what you want.  I think on the archive page you can also choose to assign a different session label to resolve the conflict.  You definitely don't want to merge, as that will put all of the data into a single session.


-Dan

gaganjyoti sharma

unread,
Jun 26, 2015, 12:21:43 AM6/26/15
to xnat_di...@googlegroups.com, gagan...@gmail.com
Thanks Daniel,

But,  Patient ID never changes throughout the study.... And they have to be same.
Normally its the studyinstanceUID and/or seriesintancesUID I have used in the past to segregate my data

Regards
Gagan 

Dan Marcus

unread,
Jun 26, 2015, 12:44:25 AM6/26/15
to xnat_di...@googlegroups.com
Right patient ID stays the same within study but should change between studies. Xnat is recognizing that they study uids are different which is why the studies are separated in the prearc. But it is giving you a conflict on archiving bc it thinks all of the session labels (extracted from patient id) are the same. 

Sent from my iPhone
--
You received this message because you are subscribed to the Google Groups "xnat_discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to xnat_discussi...@googlegroups.com.
To post to this group, send email to xnat_di...@googlegroups.com.
Visit this group at http://groups.google.com/group/xnat_discussion.
For more options, visit https://groups.google.com/d/optout.

Gagan Sharma

unread,
Jun 26, 2015, 12:59:51 AM6/26/15
to xnat_di...@googlegroups.com
Sorry Dan,

For our subjects inter and intra ID remains the same. They never change and if they do, we map them.....
In a pool of thousand scans three time points, if I have each time a new PatientID, its impossible to view them in a single folder.





--
You received this message because you are subscribed to a topic in the Google Groups "xnat_discussion" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/xnat_discussion/myJVKXgqNjk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to xnat_discussi...@googlegroups.com.

Mohana Ramaratnam

unread,
Jun 26, 2015, 1:05:03 AM6/26/15
to xnat_di...@googlegroups.com
Could you organize them as SUBJECTID_v001, SUBJECTID_v002 etc?

Where v*** indicates the visit number.

Mohana

Marcus, Dan

unread,
Jun 26, 2015, 1:05:20 AM6/26/15
to <xnat_discussion@googlegroups.com>
Not sure I totally follow but if you're stuck with whatever is in patient id then you can use the study comments field instead. See link in thread below for the format. Or you can edit the session label in the form when you archive (though that doesn't work well if you are doing large batches in auto archive mode). 

Sent from my iPhone

Gagan Sharma

unread,
Jun 26, 2015, 1:24:13 AM6/26/15
to xnat_discussion
Thanks guys but,


We normally sort our folders in this way:

 ID+StudyDate"T"Studytime+'/'+SeriesNumber_SeriesName+SeriesInstanceUID

 which gives us a nice structure like below:

P0059_01:
20140619T155722  20140619T180916  20140620T152123

P0060_04:
20140626T112737  20140626T204650  20140627T143745

P0061_05:
20140707T110044  20140708T112159

P0062_29:
20140711T210518  20140712T180804

P0063_29:
20140715T114924  20140716T130956

That is what I want to achieve...... Is there a file called dicom-import-context.xml (it exist in three different location on my virtual machine), I can edit or please let me know some other file.

Regards
Gagan Sharma


Regards
Gagan

Daniel Marcus

unread,
Jun 26, 2015, 11:19:40 AM6/26/15
to xnat_di...@googlegroups.com, gagan...@gmail.com
Gagan,

It's a bit of an effort to modify the DICOM import logic to achieve the kind of custom mapping you're looking for, but dicom-import-context is the file you'd need to work with. I suspect you'll be able to rig it up to do what you want. General instructions here: https://wiki.xnat.org/display/XNAT16/DICOM+identifier+configuration+and+customization.  If Kevin is lurking, he might have some additional guidance for you.

-Dan
To unsubscribe from this group and stop receiving emails from it, send an email to xnat_discussion+unsubscribe@googlegroups.com.
To post to this group, send email to xnat_discussion@googlegroups.com.

--
You received this message because you are subscribed to a topic in the Google Groups "xnat_discussion" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/xnat_discussion/myJVKXgqNjk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to xnat_discussion+unsubscribe@googlegroups.com.
To post to this group, send email to xnat_discussion@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "xnat_discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to xnat_discussion+unsubscribe@googlegroups.com.
To post to this group, send email to xnat_discussion@googlegroups.com.

--
You received this message because you are subscribed to a topic in the Google Groups "xnat_discussion" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/xnat_discussion/myJVKXgqNjk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to xnat_discussion+unsubscribe@googlegroups.com.
To post to this group, send email to xnat_discussion@googlegroups.com.

Gagan Sharma

unread,
Jun 28, 2015, 3:48:00 AM6/28/15
to xnat_discussion
Thanks Dan....

I will tomorrow....

Regards
Gagan Sharma

-Dan
To unsubscribe from this group and stop receiving emails from it, send an email to xnat_discussi...@googlegroups.com.
To post to this group, send email to xnat_di...@googlegroups.com.

--
You received this message because you are subscribed to a topic in the Google Groups "xnat_discussion" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/xnat_discussion/myJVKXgqNjk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to xnat_discussi...@googlegroups.com.
To post to this group, send email to xnat_di...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "xnat_discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to xnat_discussi...@googlegroups.com.
To post to this group, send email to xnat_di...@googlegroups.com.

--
You received this message because you are subscribed to a topic in the Google Groups "xnat_discussion" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/xnat_discussion/myJVKXgqNjk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to xnat_discussi...@googlegroups.com.
To post to this group, send email to xnat_di...@googlegroups.com.

--
You received this message because you are subscribed to a topic in the Google Groups "xnat_discussion" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/xnat_discussion/myJVKXgqNjk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to xnat_discussi...@googlegroups.com.
To post to this group, send email to xnat_di...@googlegroups.com.

goo...@merctech.com

unread,
Jul 8, 2015, 3:56:52 PM7/8/15
to xnat_di...@googlegroups.com
In the message dated: Fri, 26 Jun 2015 05:05:10 -0000,
The pithy ruminations from "Marcus, Dan" on
<Re: [XNAT Discussion] Re: Multi time point patient data.> were:
=> Not sure I totally follow but if you're stuck with whatever is in

We've got the same issue -- clinical scanners here are configured to push
cases automatically to our instance of XNAT. These scans may be follow-ups
for existing patients, where an earlier scan has been entered into XNAT,
so we frequently have cases declared as a "Conflict" and not archived.

A while ago I tried to address the problem by re-writing the Session_ID for
these conflicting cases in the prearchive so that they'd be archived under a
new (non-conflicting) name, but I ran into difficulty with the REST syntax:

https://groups.google.com/forum/#!topic/xnat_discussion/kI0KUMIuimg

any help would be appreciated.

=> patient id then you can use the study comments field instead. See link
=> in thread below for the format. Or you can edit the session label in the
=> form when you archive (though that doesn't work well if you are doing
=> large batches in auto archive mode).

In our situation, we really cannot ask the clinical radiology technicians
to add or edit DICOM fields before cases are pushed.

The general question of resolving conflicting subject IDs or re-naming
subjects seems to come up fairly often, and it would be terrific to have
a way for XNAT users to easily specify the unique key used within the
database (ie., SubjectID+timestamp). By default, this could remain the
subject ID from the DICOM data, preserving the current behavior.

Thanks,

Mark

=>
=> Sent from my iPhone
=>
=> On Jun 25, 2015, at 11:44 PM, Dan Marcus
=> <dan.m...@gmail.com<mailto:dan.m...@gmail.com>> wrote:
=>
=> Right patient ID stays the same within study but should change between
=> studies. Xnat is recognizing that they study uids are different which
=> is why the studies are separated in the prearc. But it is giving you a
=> conflict on archiving bc it thinks all of the session labels (extracted
=> from patient id) are the same.
=>
=> Sent from my iPhone
=>
=> On Jun 25, 2015, at 11:21 PM, gaganjyoti sharma
=> <gagan...@gmail.com<mailto:gagan...@gmail.com>> wrote:
=>
=> Thanks Daniel,
=>
=> But, Patient ID never changes throughout the study.... And they have to
=> be same. Normally its the studyinstanceUID and/or seriesintancesUID I
=> have used in the past to segregate my data
=>
=> Regards Gagan
=>
=> On Friday, 26 June 2015 14:06:07 UTC+10, Daniel Marcus wrote: Hi Gagan,
=>
=> XNAT is designed to support the sort of data representation you're
=> looking for. If you're getting a conflict on archiving, I suspect maybe
=> you have the same value in the DICOM patient ID field for all of your
=> sessions. That's the field that XNAT uses to distinguish individual
=> sessions. Try putting different values in that field for the different
=> DICOM studies and it will probably do what you want. I think on the
=> archive page you can also choose to assign a different session label to
=> resolve the conflict. You definitely don't want to merge, as that will
=> put all of the data into a single session.
=>
=> https://wiki.xnat.org/display/XNAT16/Basic+DICOM+object+identification
=>
=> -Dan
=>
=> On Thursday, June 25, 2015 at 10:34:51 PM UTC-5, gaganjyoti sharma
=> wrote: Hi all,
=>
=> Please correct me if I am wrong.
=>
=> In our lab we have lot of longitudinal data, multiple time points
=> I(MR/CT) for the same patient. tried to push that data set towards XNAT.
=> INow it was well received in pre archive, but archiving gave me conflict
=> Iand when I tried to merge manually, the output was not good/favourable.
=> Ideally we want patient data archived in ID/Date_time/series format. Is
=> Iit possible or I need to write a special pipeline for that ?
=>
=> Regards Gagan Sharma
=>
=> PS: Any good step by step tutorial for pipeline, please post.
=>
=> -- You received this message because you are subscribed to the
=> Google Groups "xnat_discussion" group. To unsubscribe from
=> this group and stop receiving emails from it, send an email to
=> xnat_discussi...@googlegroups.com<mailto:xnat_discussion+unsu
=> bsc...@googlegroups.com>. To post to this group, send email to
=> xnat_di...@googlegroups.com<mailto:xnat_di...@googlegroups.com
=> >. Visit this group at http://groups.google.com/group/xnat_discussion.
=> For more options, visit https://groups.google.com/d/optout.
=>
=> -- You received this message because you are subscribed to the
=> Google Groups "xnat_discussion" group. To unsubscribe from
=> this group and stop receiving emails from it, send an email to
=> xnat_discussi...@googlegroups.com<mailto:xnat_discussion+unsu
=> bsc...@googlegroups.com>. To post to this group, send email to
=> xnat_di...@googlegroups.com<mailto:xnat_di...@googlegroups.com
=> >. Visit this group at http://groups.google.com/group/xnat_discussion.
=> For more options, visit https://groups.google.com/d/optout.
=>
=> -- You received this message because you are subscribed to the
=> Google Groups "xnat_discussion" group. To unsubscribe from
=> this group and stop receiving emails from it, send an email to
=> xnat_discussi...@googlegroups.com. To post to this group,
=> send email to xnat_di...@googlegroups.com. Visit this group at
=> http://groups.google.com/group/xnat_discussion. For more options, visit
=> https://groups.google.com/d/optout.
=>
--
Mark Bergman Biker, Rock Climber, Unix mechanic, IATSE #1 Stagehand
'94 Yamaha GTS1000A^2
ber...@panix.com

http://wwwkeys.pgp.net:11371/pks/lookup?op=get&search=bergman%40panix.com

I want a newsgroup with a infinite S/N ratio! Now taking CFV on:
rec.motorcycles.stagehands.pet-bird-owners.pinballers.unix-supporters
15+ So Far--Want to join? Check out: http://www.panix.com/~bergman

Felix Ruiz

unread,
Aug 18, 2015, 3:05:28 PM8/18/15
to xnat_discussion, goo...@merctech.com
Reply all
Reply to author
Forward
0 new messages