dcm4chee-web3 displays some studies under wrong patient

[Björn Albers]

Hey,

we intentionally have many individuals with two patient records within our dcm4chee (v.2.17.1): one record with a "normal" patient ID and one with a patient ID from our mammography screening software (both have to coexist in our workflow).

The web-gui (dcm4chee-web3) sometimes displays studies under the wrong patient and it's not possible to fix this by moving the studies.

An example with anonymized patient...

Given the following patient records with equal name and birth date but different patient ID:

mysql> select pk,pat_id,pat_name,pat_birthdate from patient where pat_name like 'WINTER^IRIS%';
+--------+-----------------------------------+----------------+---------------+
| pk     | pat_id                            | pat_name       | pat_birthdate |
+--------+-----------------------------------+----------------+---------------+
|  54851 | 23773                             | WINTER^IRIS^^^ | 19620528      |
| 132213 | 1.2.276.0.85.049.20.04.0002342443 | WINTER^IRIS^^^ | 19620528      |
+--------+-----------------------------------+----------------+---------------+
2 rows in set (0.00 sec

The patient with regular patient ID has these studies:

mysql> select pk,study_id,study_datetime,mods_in_study from study where patient_fk = '54851' order by study_datetime;                                                                                                                            
+--------+------------+---------------------+---------------+
| pk     | study_id   | study_datetime      | mods_in_study |
+--------+------------+---------------------+---------------+
| 150707 | 56122568   | 2006-05-02 14:20:26 | MR            |
|  84775 | 0000027828 | 2007-04-25 23:24:12 | CT            |
|  79290 | 230916366  | 2007-04-27 15:26:06 | PR\MR\SR      |
| 162957 | NULL       | 2007-12-17 16:09:12 | CR            |
| 186428 | 365720     | 2010-09-20 16:07:12 | MG            |
| 256579 | 413686     | 2012-10-23 08:52:21 | MG            |
| 294748 | 1          | 2012-10-23 09:23:51 | US            |
| 292576 | 29572      | 2013-11-12 15:10:41 | CT\SR         |
| 294592 | 441042     | 2013-11-29 10:14:25 | PR\MR         |
| 294721 | 441192     | 2013-12-02 13:48:30 | PR\MR\SR      |
+--------+------------+---------------------+---------------+
10 rows in set (0.00 sec)

...and the second patient (screening ID) has only one study:

mysql> select pk,study_id,study_datetime,mods_in_study from study where patient_fk = '132213' order by study_datetime;
+--------+------------+---------------------+---------------+
| pk     | study_id   | study_datetime      | mods_in_study |
+--------+------------+---------------------+---------------+
| 255236 | 0002342443 | 2012-10-10 08:34:42 | MG            |
+--------+------------+---------------------+---------------+
1 row in set (0.00 sec)

When I search in dcm4chee the patient then some studies get displayed unter the wrong patient (i.e. the US-study should be listed unter patient ID 23773 instead 1.2.276.*, see screenshot), even though the dicom files include the right patient ID (retrieved and checked via OsiriX):

PatientsName (0010,0010) Winter^Iris
PatientID (0010,0020) 23773
PatientsBirthDate (0010,0030) 19620528

Did I found a bug?

Björn

PS: This is a follow-up post of https://groups.google.com/forum/#!topic/dcm4che/3NbGfbV6UU8

[Björn Albers]

The version of dcm4chee-web is 3.0.1 (16430).

[Björn Albers]

I found out the studies are only listed under the wrong patient if I search by patient name.

When I search by patient id then the studies appear under the right patient (see attached screenshots).

So there must by a bug in the filter or somewhere nearby (tried to look-up the bug in the source but I guess I won't be helpful here, sorry... java-noop).

[fleetwoodfc]

Can you add the merge_fk to your patient query i.e. select pk,pat_id,pat_name,pat_birthdate, merge_fk from patient where pat_name like 'WINTER^IRIS%';

[Björn Albers]

On Wednesday, December 4, 2013 6:49:13 PM UTC+1, fleetwoodfc wrote:

Can you add the merge_fk to your patient query i.e. select pk,pat_id,pat_name,pat_birthdate, merge_fk from patient where pat_name like 'WINTER^IRIS%';

Both records have NULL as merge_fk.