Posts by bigg

  • Images arriving out of sequence

    It seems as though our AGFA CR systems are sending images out of sequence to our Conquest archive.
    The problem is that the CR's are sending over slow WAN links for teleradiology and there is no way for the doctors to tell how many images should be in the study. The radiographers now annotate on the 1st image in a series, but the images are arriving out of sequence which is upsetting the doctors who have to constantly open and close a study to see if the first image has arrived.


    Is there any way to have a study available (or visible in the db) only once the whole study is received by conquest? OR is there a way to force the images to be received in order?
    (Agfa says that their systems can not be modified to send images in the correct order.)


    Is there a RecieveAssocian level setting somewhere?


    Gavin

  • Multi Threading?

    Thanks guys.


    I am busy testing forward study for now. Will see how it goes.
    Any advice on tweaking mySQL to up performance?


    Gavin

  • Multi Threading?

    I was looking in the manual again :)


    I saw some things relating to the forward association level and I was wondering if I changed the forward association level to Patient if it would increase write performance.
    As far as I can determine what is happening currently is the following:
    1.) Image Recieved
    2.) Image converted to j2
    3.) Image converted to un
    4.) Image forwarded to Server
    5.) Goto 1.


    Would it make any performance difference if I the forward and import converter instead of an export converter?
    Or would it be better to change the forward association to patient?


    Gavin

  • Multi Threading?

    The server itself is not slow, but when sending images to Conquest it is slow.
    I do not see anything obvious in the task manager that shows the server is taking strain.


    We have an auto-forward enabled:
    ForwardAssociationLevel = IMAGE
    ExportConverters = 1
    ExportConverter0 = forward to IMPAX62


    I am storing the images as J2.
    DGATE.exe is only using about 60MB of memory at the moment
    mysql is only using about 31MB of Memory
    The database has +/- 3.2 Million images in the images database.


    I tried increasing the QueueSize from 128 to 256 and have noticed no difference.
    Any other ideas on what could be causing slow write speeds?

  • Multi Threading?

    Is there a way to improve the performance of this queue?


    EDIT:
    I was looking through the documentation and noted the following:

    Quote

    The following hidden option exists: when the service buttons are alt-right clicked, the service is
    installed four times (e.g., with ports 5678…5681). Each server runs independently against the
    same data(base). Use to increase speed or reliability or for testing purposes.


    I was wondering how this works? Does it install 4 seperate instances that need to be addressed indipendatly?
    (i.e. will I have to configure the modalities to access the server on different ports to increase performance?)

  • Multi Threading?

    Hi Marcel,


    Is Conquest multi-threaded? Does it allow for more than one process to happen at the same time or are commands processed sequentially?
    We are having some speed issues and we can not identify the cause.
    We have a Windows 2008 Server x64 with Intel Xeon 5405 2GHz, 4GB RAM.
    The system does not seem to be taking any strain and we do not appear to have any networking issues.
    We suspect that Conquest may be queuing tasks which is resulting in the slowdown.
    Is this possible or is there something we are missing?


    Regards,
    Gavin

  • Philips MRI Problem

    Hi Marcel,


    I am having a problem with images from our Philips MRI.
    Brief run down of the problem:
    When sending images to our one conquest archive we can not retrieve the images from our philips ViewForum workstation or any other workstation (K-Pacs, Centricity or ClearCanvas).
    I noticed that if I disable
    1.2.840.10008.5.1.4.1.1.11.1
    1.2.840.10008.5.1.4.1.1.4.1
    Then i can send and receive fine from any modality, I just can't access any old cases that have these SOP Classes (Which is a huge problem)
    If I send images from the same modality to our other conquest server everything is fine. This server does not have these SOP Classes disabled.


    As far as I can determine both Conquest servers are running the same versions of Conquest and have mostly identicle settings.
    Both are running on 64bit (One Windows 2008 and the Other Vista), both have jpeg lossless compression enabled.


    Is there any other setting that may be causing the one Conquest server to not send these images?


    EDIT:
    Both servers running version 1.4.14
    Found that there is a difference in the dgatesop.lst
    The server having a problem has "LittleEndianExplicit 1.2.840.10008.1.2.1 transfer" disabled.
    Could this be the cause of the problem? Will it impact on anything else if I enable it?

  • Multiple Scheduled Workstations in Worklist?

    Hi Marcel,


    I am setting up worklist on Conquest and I need to be able to have multiple AETitles in the ScheduledA field in the database.
    I was thinking of trying to create duplicate entries in the worklist table, but I can't have 2 rows with the same accession number due to key constraints.
    Is there any other method to place 2 AETitles in the ScheduledA field so that 2 ultrasound modalities can be assigned to the same worklist item?

  • acrnema.map comments

    Hi Marcel,


    I have picked up some issues when commenting in the acrnema.map file.
    Are there any limitations in this regard?
    EXAMPLE:

    Code
    /* THIS IS A TITLE COMMENT */

    AETITLE1     192.168.1.1   un   /* this is a comment */

    AETITLE2     192.168.1.2   un   # this is a comment

    AETITLE3     192.168.1.3   un   // this is a comment

    AETITLE4     192.168.1.4   un   # this is a comment that wraps to the next line when viewed in the conquest GUI.


    In the example above I noticed that AETITLE1 does not display in the DICOM System or Destination drop down lists in the GUI.
    AETITLE2 & AETITLE3 display fine.
    AETITLE4 results in AETITLE4 and the word "the" being displayed in the destination and dicom system dropdowns.


    What is the limit on the length of a line in the acrnema.map file? And is there a commenting method that can prevent these wraparounds from causing havoc?

  • *** multiplex: connection terminated

    Hi Marcel,


    Bad news. It would seem that the error is both for the Dude traffic as well as for Modality Worklist.
    Our AGFA CR could not query the worklist since the upgrade. (The radiographers only noticed this morning)
    I don't know if anyone else has experience this problem yet?


    Gavin

  • *** multiplex: connection terminated

    Solved!


    Hey Marcel, I think I found out what is causing the problem.
    I setup The Dude (http://www.mikrotik.com/thedude.php) to monitor that server and it polls it on port 104 every 30 seconds.
    It seems that this is what is causing the error to come up as it is not a valid DICOM communication but merely probes the port to check that the service is available.
    Is there any way to disable this message? it is filling up the server status screen.


    Gavin

  • *** multiplex: connection terminated

    Hey Marcel,


    I recently wrote an application that takes information from our RIS system and adds it directly to the Conquest Worklist database.
    There have been problems with some modalities such as Philips MR and some ultrasound machines.
    Our Agfa CR and Fuji CR seem to work fine.
    I think it may be due to a DIcom conformance issue with the accession number or study instance id as I am using our account number from the RIS for these.
    Our account number is in the format AA000000. I think some modalities are looking for a more standard AccessionNumber/study instance id.
    I wonder if this could be causing the error?


    Gavin

  • forward studies - start and stop with batch ?

    Easier solution than my last one:


    create two dicom.ini files one for during the day and one for after hours.
    Use xcopy to overwrite the dicom.ini file in your conquest folder.
    The other way I had before needed way too many lines of code to actually do something.
    Thanks to my work colleague for pointing out my stupidity.

  • forward studies - start and stop with batch ?

    Hey Uli,


    I am not sure if this will work as I have not tried it, but maybe you could write a script that re-writes the dicom.ini file at certain times of day.
    i.e. The script adds and removes forwarding rules from your dicom.ini file. You can the setup the script to remove forward entries at night when the machines are off and add them again in the morning when the computer should be back on again.


    Its a bit of a dirty solution, but it may work.
    Let me know how it goes.


    what you will have to do is create two batch scripts for each time you want to change the dicom.ini file.
    One file to echo the desired dicom.ini output and one to pipe it into a text file.


    i.e.


    Batch file 1 : (bat1.bat) Overwrite the existing dicom.ini file

    Code
    @echo offcall bat2.bat > dicom.ini@echo on


    Batch file 2 : (bat2.bat) Echo the dicom.ini content

    Code
    echo [sscscp]

    echo MicroPACS                = sscscp

    echo Edition                  = Personal

    ......

    ......

    All your config etc.