Extract all images from DCM4CHEE database for recovery purposes

[Richard Lawrence]

Hi

Afraid this might be another dumb question form one with just enough knowledge to be dangerous - anyways here it is :-

Our vet practice has grown over last 7 yrs from one CR unit and DVD backup to 4 sites 12 modality units and multiple CONQUEST severs - one at each of 4 sites and a main backup one at one of the sites - (our equine clinic was the obvious place because they generate about 90% of our images) Obviously this site has its own primary server and the backup server so doesn't have any off site backup

I use the CONQUESTs because they are what I cut my teeth on 7 yrs ago and they were very easy to copy images off to a USB hard drive for extra backup safety - just copy the data folder and there you are - HD failure? - no problemo - just copy back saved images and reinitialize database.

But now I have set up a DCM4CHEE server based in town to backup the daily images from the Equine site. I really like the DCM4CHEE option because in  opens up the ability to allow referring vets to see their case images via weasis utilizing the permissions function that the kind members of this group helped me configure. I also have DCM4CHEE installed on the main CONQUEST backup server for in house oviyam use.

Finally the question:

If my Equine CONQUEST and the group backup are lost in a catastrophic event (bush fire is a real possibility here - going to hit 41C on Sat!!)  I still have Equine images backed up on the remote DCM4CHEE machine (but now that is the only copy I have) - so all is well for the time being, but can I somehow drag/extract all the DICOM folder/images off that machine and load them back onto a new/rebuilt CONQUEST?. I know where the DCM4CHEE archive is obviously, but after that I am stuck

At the moment I am forwarding all images as they come into DCM4CHEE to an old box with CONQUEST on it and that way I can access the DICOM folders easily but its clumsy. The other options I have considered are that if I have a failure and DCM4CHEE is my only backup then I can just make a clone of the DCM4CHEE machine or I could send all images from the DCM4CHEE backup back to a new CONQUEST machine but that would take  a long time I suspect.

Or maybe I'm doing this all wrong?

Any suggestion welcome

Cheers Richard

[fleetwoodfc]

but can I somehow drag/extract all the DICOM folder/images off that machine and load them back onto a new/rebuilt CONQUEST?.

I think so because DCM4CHEE stores standard DICOM files - but try it!

Notes: By default DCM4CHEE stores a copy of the original DICOM file and any edits made update the database and not the file. So if you rebuild from the files only then updates will be lost.  

 If you send from DCM4CHEE then updates are included.

Have you considered cloud storage or even running your DCM4CHEE server in the cloud?

You might also consider using OSIRIX (MAC based) as it basically does everything that CONQUEST can plus it has a built-in web portal that includes Weasis and user access management. 

[Richard Lawrence]

Sorry - been away for a while

Thanks for the post - I have looked at Osirix before but lack of Macs at work is the limiting factor and I am very happy with most of our current system

Cloud is too expensive and slow for us - Our internet speeds are pretty slow.

I don't think DCM4CHEE store images as .dcm files - at least when I search the archive folder there are none there - just a whole load of meaningless (to me file and folder names)

Does anyone know if you can extract a dicom image/file directly from the DCM4CHEE archive directory

Thanks Richard

[Dr Suresh Viswanathan]

dcm4chee stores all the files as standard DICOM files (depending on your configuration, the files might be compressed). As per the DICOM standard, files shouldn’t have a meeningful name nor should they have a .dcm extension. However the files you see can be opened in any standard DICOM viewer.

Suresh.

--
You received this message because you are subscribed to the Google Groups "dcm4che" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dcm4che+u...@googlegroups.com.
To post to this group, send email to dcm...@googlegroups.com.
Visit this group at https://groups.google.com/group/dcm4che.
For more options, visit https://groups.google.com/d/optout.

[Algie Rosario]

You dont have to extract the dicom files in the DCM4CHEE Archive Directory, it is the standard Dicom Images. You can open it with Weasis, Efilm, Osirix, or any standard dicom viewer. Have tried it multiple times bro.

[Algie Rosario]

[Algie Rosario]

[Richard Lawrence]

Thanks to all for their replies but I don't think I have explained myself properly and/or there may not be a suitable answer

I have no problem retrieving dicom images from dcm4chee archive using any of our numerous viewing platforms

My issue is that with a CONQUEST archive all the dcm files are stored in the archive folders and the whole archive can be copied and transferred to a new instance of CONQUEST and then the database regenerated, and all is recovered

If however I totally loose my CONQUEST archive how do I regenerate it from a backup DCM4CHEE storage archive. Sure I could Q/R every one of our 2TB of images from DCM4CHEE to CONQUEST but I was interested to know if there is another way to copy the dicom archive from DCM4CHEE to CONQUEST (or even DCM4CHEE to DCM4CHEE if you loose the database and only have the archive?)

Hope that makes sense?

Richard

On Wednesday, 16 December 2015 11:09:51 UTC+11, Richard Lawrence wrote:

[Sinan Sahin]

You can use storescu command with parameters like

storescu -c AEOFPACS@IPOFPACS:port folder to be sent

storescu -c DCM4...@192.168.3.205:11112 X:\DICOM

With this all the images will be pushed to the pacs

I hope it may help.

--

[Richard Lawrence]

Thanks some1

I will have a look at that but excuse the ignorance but where do I run the storescu command from? 

The command prompt in the DCM4CHEE directory? or somewhere else

thanks Richard

On Wednesday, 16 December 2015 11:09:51 UTC+11, Richard Lawrence wrote:

[Sinan Sahin]

You need to install dch4che toolkit 3, its in the bin subdirectory. You need to run it at command prompt 

Hope will help.

[Richard Lawrence]

Many thanks - will have a play and see what I can do

On Wednesday, 16 December 2015 11:09:51 UTC+11, Richard Lawrence wrote:

[matthe...@netscape.net]

So, I have not actually done this for a long time but if I remember correctly...

Copy $dcm4cheeHome/server/default/archive directory to a USB drive.

Disconnect the USB drive from the dcm4chee server.  Connect to PC with Conquest.
Make sure your new Conquest install is set to the same storage syntax (RAW, JPEG2k or JPEG) as the data copied from the dcm4chee server.

Open conquest.  Make sure drag n drop (I cannot remember the exact name) =1  (dicom.ini file)
This allows dragging and dropping a folder onto conquest.

Conquest will copy in all the data from the dropped folder.

--- you "CAN" use a different storage syntax, but it will take alot longer since Conquest will recompress to the configured storage scheme.
It also stuffs all the dicom into the conquest database.

For 2tb, this might take awhile, but once it finishes, you are done.  You can also accept new imaging over th network while the import is happening.

There is a "trick" for conquest also:

On the Configuration interface there is a radio button for "Install as a service" or "run as a service"  ... something like that..

Hold down the shift key and select that radio button.  This runs multiple (4?) instances of the Conquest server.

Anyway... that is the process.. setup the new conquest server and configure it.
Drag and drop the folder containing all the dicom onto the Conquest interface.
Come back sometime later when it is all done.

Hope this helps.

[Richard Lawrence]

Hi guys - Just a quick thank you for your tips and adv

As always seems to happen I am nearly at my goal but there are a last few headaches

Using the storescu from dcmtoolkit 3 returned and  "error: unknown option -c"

So not sure what's going on there -When I have time I shall investigate the syntax for storescu but it looks like that might do the job

Using Matthe..'s suggestion of drag and drop into Conquest works brilliantly ..... to a point

Firstly I could find no reference to drag and drop in the dicom.ini file but that didn't matter because drag and drop works off the bat anyway

However when I set up the DCM4CHEE I chose the JPEG-LS_Lossless option which doesn't seem to be supported in Conquest - I get all decompression errors, full database with image details but no images..aargh

So when I have time I plan to change the DCM4CHEE to save uncompressed and then try again. So far I don't need to extract images so I have time to play around. I'm just trying to prepare for disaster

Thanks again for all the suggestions Richard

**STOP PRESS**  I tried sending one image and changed DCM4CHEE to store as Lossless JPG and then did an archive drop on a fresh install of Conquest configured to Lossless JPEG and it worked all crap except the one new image which I can view fine.. Yeah. Shall do a few more tests but This may have cracked it

Cheers Richard

On Wednesday, 16 December 2015 11:09:51 UTC+11, Richard Lawrence wrote:

[matthe...@netscape.net]

ImportExportDragAndDrop  = 1

IF there is any errors related to trying to drag and drop a 2tb folder...  drop in the subfolders separately.

On install of Conquest, on the first tab of the Conquest server or in the dicom.ini file, you can set compression: 

 

On the server interface, select the "enable Jpeg (2000) checkbox.

This allows selection of the various jpeg options under "Images are stored on disk as:" 

Make a selection

In the dicom.ini file  (remember to stop the server, edit, save, then restart the server)

# Configuration of compression for incoming images and archival

DroppedFileCompression   = jl    (as titled   .probably should have used j1 here on your example.. though could have been j2.. see compression code list)

IncomingCompression      = jl      (for association)

ArchiveCompression       = as      (Store 'as' received / dropped.  Change to desired archival compression)

For compression see manual. Values are 

un=uncompressed;  

 ul=littleendianexplicit,

ub=bigendianexplicit,

ue=both     

j1,j2=lossless jpeg;j3..j6=lossy jpeg;

n1..n4=nki private 

jk   =lossless jpeg2000;jl=lossy jpeg2000 

Use these codes in known dicom providers also...  and list the local conquest AEtitle  127.0.0.1 port 11112

 

[Richard Lawrence]

Thanks... now I see the drag and drop reference. Interestingly ImportExportDragandDrop is set to default of = 0 but still works OK - Anyway I have set it to = 1 to be safe

In the interim I changed the default storage compression on my DCM4CHEE to JPLL = Jpeg lossless as this is one of the types of compression supported by Conquest

I didn't have to do anything else, just start a fresh install of CONQUEST enabling JPEG(2000) support and choosing Lossless JPEG.

I also deleted all images from DCM4CHEE archive (fortunately not that many) and resent them to be stored in correct format

On testing I can drag and drop the DCM4CHEE archive into CONQUEST and after a little time I have a new CONQUEST database (no errors) with all the images nicely stored as .dcm files (rather than the esoteric DCM4CHEE storage format) and from there I can do what I like with them

Which is exactly the ability I wanted incase my CONQUEST failed and all i had left was the DCM4CHEE archive as a backup

So thanks to all

Of note I didn't have to add the extra config lines as suggested by matthe, but maybe I was lucky

I'm not sure if this would have worked if I had kept the original storage choice of  JLSL - JPEG-LS Lossless that I made in DCM4CHEE

I tried all the compression choices that one can select from the CONQUEST Configuration page and none of them worked on importing JPEG-LS lossless files, it might be that one of the others mentioned 

by matthe and added to the dicom.ini may have worked but since everything is OK as I  now have it I haven't explored any further

Once again many thanks to all that contributed especially matthe...

On Wednesday, 16 December 2015 11:09:51 UTC+11, Richard Lawrence wrote: