Inconsistency in database
98 views
Skip to first unread message
Cole williams
Jul 7, 2020, 10:26:19 AM7/7/20
to dcm4che
Hello,
Recently I imported a patient folder into a PACS that runs on the 5.22.1 secure-ui version of dcm4chee5-arc-light.
Recently I imported a patient folder into a PACS that runs on the 5.22.1 secure-ui version of dcm4chee5-arc-light.
I found that this patient was incoherent at the BdD level. Indeed, as you can see on the screenshot, we notice that for the same patient we have two records, the first with all the patient's studies and the second that contains only one study.
The inconsistency does not end there since when you roll/unroll at the patient tree, you see that at that time it integrates the 10 patient exams into the two records listed just before.
To go further, I also note that when a search is made on the day of the examination date (24/03/2020), only the SR is found in patient file number 1 but no trace of the XA and OT is found in the second patient file.
If you do a search by the Patient ID or the Patient Name, the XA and OT exam will appear.
I would like to understand what is going on so that I can solve this problem of consistency in the ACAP search engine. Is this due to inconsistency in the database? A concern in my patient record that I imported?
In advance, thank you for your help,
Cordially,
Cole
Shefki Esadi
Jul 7, 2020, 12:00:10 PM7/7/20
to dcm4che
Hi,
Regarding to first part:
- First of all you are in the Study Tab, that is a list of studies not of Patients, so in every study that is coming back from the request there are the patient attributes in there too. All studies in the order as they come, that have the same patient attributes are collected in one Patient, if at least one attribute of the patient (of the next study in the list) is different (besides the `(0008,0005): Specific Character Set`) that Study will be collected in another patient. So check which attribute is has differed and why.
Regarding to second part:
We have the function in-there if you close the tree and open again, the UI will query all studies to that patient (with that patient ID). (I know that, that can be confusing, that comes from the Design as the UI study page was all in one, we had a discussion about that if that is a good idea to get all the studies when you close and open the tree). Because the request is made for a specific Patient ID, the compare of patient attributes of the requested studies is not made anymore so you have all 10 studies shown to the same patient. (We can discus internally that feature again maybe is a good idea to remove that)
The third part is not much clear for me. Maybe you can open the browser Inspect (on the most browsers by clicking F12) go to the Network tab, and see what query is made there. I guess the first results is querying with the date filter and second is getting only with patient ID and that's why is getting everything what that patient has even the studies with XA and OT. (Maybe it is a bug that I should fix).
Best Regards
Shefki Esadi
Regarding to first part:
- First of all you are in the Study Tab, that is a list of studies not of Patients, so in every study that is coming back from the request there are the patient attributes in there too. All studies in the order as they come, that have the same patient attributes are collected in one Patient, if at least one attribute of the patient (of the next study in the list) is different (besides the `(0008,0005): Specific Character Set`) that Study will be collected in another patient. So check which attribute is has differed and why.
Regarding to second part:
We have the function in-there if you close the tree and open again, the UI will query all studies to that patient (with that patient ID). (I know that, that can be confusing, that comes from the Design as the UI study page was all in one, we had a discussion about that if that is a good idea to get all the studies when you close and open the tree). Because the request is made for a specific Patient ID, the compare of patient attributes of the requested studies is not made anymore so you have all 10 studies shown to the same patient. (We can discus internally that feature again maybe is a good idea to remove that)
The third part is not much clear for me. Maybe you can open the browser Inspect (on the most browsers by clicking F12) go to the Network tab, and see what query is made there. I guess the first results is querying with the date filter and second is getting only with patient ID and that's why is getting everything what that patient has even the studies with XA and OT. (Maybe it is a bug that I should fix).
Best Regards
Shefki Esadi
Cole williams
Jul 9, 2020, 3:57:52 AM7/9/20
to dcm4che
Hi,
First of all, thank you for your return. With regard to the first point, that's what I immediately thought, that there is a difference in one of the attributes of the patient but unfortunately after several checks no difference has been found.
With respect to the last and third point, I would like to point out that the Q/R was done from several different console systems and the same result was observed.This was not seen only from the UI Web.
Best Regards
Cole
First of all, thank you for your return. With regard to the first point, that's what I immediately thought, that there is a difference in one of the attributes of the patient but unfortunately after several checks no difference has been found.
With respect to the last and third point, I would like to point out that the Q/R was done from several different console systems and the same result was observed.This was not seen only from the UI Web.
Best Regards
Cole
Shefki Esadi
Jul 9, 2020, 6:16:49 AM7/9/20
to dcm4che
Hi,
to be able to debug the problem, maybe you can send me the request (at least the study 9 and 10) ( you can remove/anonymize the attributes that are patient sensitive, even the whole study relevant attributes you can remove).
Best Regards
Shefki Esadi
to be able to debug the problem, maybe you can send me the request (at least the study 9 and 10) ( you can remove/anonymize the attributes that are patient sensitive, even the whole study relevant attributes you can remove).
Best Regards
Shefki Esadi
gunterze
Jul 10, 2020, 7:11:45 AM7/10/20
to dcm4che
Can you provide the server.log around the time of the Patient Create Date Time of the second Patient.
Cole williams
Jul 21, 2020, 10:14:03 AM7/21/20
to dcm4che
Hello,
Thank you for coming back !
Sorry for the response time but I wanted to be sure to provide you with the right elements.So here are the studies in question, which are anonymized, but we still see the problem.
You will also find the section of the logs requested.
Cordially,
Cole
Reply all
Reply to author
Forward
0 new messages