.DCM format or DICOM Tree Format

    I am working with NDMA to obtain our archive of mammo studies they have been sent over the past 30 months prior to us setting up Conquest.


    I have 2 questions hoping someone can guide me in.


    1.) They have offered to me either straight forward DICOM files with a .DCM extension or a tree study similiar to /Study UID/Series UID/Image.


    Can both formats be brought into Conquest? Is the 'tree study' approach similiar to what Conquest uses?


    2.) There is 1.2 terabytes of uncompressed studies. What is the best method of importing these into Conquest? Do I drag them into the Conquest interface, bringing in a few thousand a day over severel months, or is there a better method?

    first of all, you ve to make a decision, if you want to store all the files uncompressed or compressed to your new archive.
    But how do you get al your files?
    If you want to store all the files uncompressed, you can sore all the dcm files to 1 folder and regenerate import them to your database with the regenerate function.
    If you want to store the all the files compressed, I would set up a temporary dicomserver to import all the files umcompressed as mentioned and move them with k-pacs to your new dicomserver.
    I transfered with this method round about 1 TB from 1 server to another, transferring the files in terms of six month.
    With dropping the files by hand, you ll get mad on 1,2 TB.

    I had a similar problem but didnt have another server to use. I DID have a single server with plenty of disk space so here is how I got them in and compressed them.


    -I installed 2 conquest servers and K-Pacs on the same box with different ports and AE's configured for each.


    -I copied the DICOM data into directory designated by MAG0 and did a regen on Conquest1.
    --At this point the data is available for use on the first server


    -I then configured Conquest2 with a different MAG0 directory and compression (I used n1)


    - I then used K-Pacs to move the data between servers. I used alphabetical but date ranges would work as well....it depends on how much data you have.


    -Once the data is moved to Conquest2 in a compressed form, I turned off Conquest1 and reconfigured Conquest2 to look like Conquest1 (AE and Port)


    -Then delete the uncompressed data...


    Hope this helps....

    Bushranger ... I am storing all data uncompressed. I am trying to understand the differences in the file structure and if there is a benefit between one or the other.


    I've noticed that within Conquest when a new study arrives from our GE mammo unit, or if I drag studies from Efilm workstations, that Conquest creates a directory with the medical record number under the data drive subdirectory as defined in the setup. That seems all very logical to me and makes it easy for me to understand the process. I've worked previously with a PACS vendor who utlized a single storage directory and incorporated a DICOMDIR file which I believe is an index to the files stored below it.


    If I use .DCM files, which if i understand correctly, will be contained in one directory, does that mean that each file is processed, and added to the database sequently? Let's say there are 2 studies with 5 images per study and let's say that the files are processed sequently by Conquest, and let's say that study 1/file 1 is processed first, and then study2/file3, and then study 1/file4, etc .. how does Conquest know to associate all 5 images per study within one reference?


    Or a simpler question might be, if I had a study with 5 images and I dropped 4 of them into the conquest interface, and then a day later dropped the 5th study in, will conquest know to associate the 5th image with the other 4 from that study? If the answer is yes then then I can understand how this will work.


    Bobomfb ... appreciate your suggestions.


    Besides importing these prior studies, I also will be replicating the entire database/files to a second server (both servers exactly the same hardware config) and once replicate the second server will be placed at one of our remote locations with real time replication.


    Ideally I'd like to load all these studies on both servers over a weekend and then move on. I'll give more thought to your suggestions


    thanks

    clabbadia,


    I am doing the exact same thing that you are going to do. I have 2 identical servers....one each at 2 different campuses. I am keeping them sync'd by using the forward feature to forward all studies from the primary server to the secondary. I am up to 9 million images and the sync seems to be working well.


    I am using MySQL to track the number of images by looking at the record numbers in the tables.


    My experience with your 4 of 5 images sent and then send the 5th.... I have watched the image count in the database and in my experience here is what happens:


    -If the study is already in the database (an thus conquest) and you resend it the image count does not go up. So conquest either overwrites the existing study or checks to see if it is there and does not really import it.


    -If the study is partially in the database and it is resent the image count only goes up by the amount of missing images.


    Hope this helps.....

Participate now!

Don’t have an account yet? Register yourself now and be a part of our community!

Register Yourself Login