Dicom changes in 1.4.19?

Greetings, we've run into a small problem with the new version.

Previously, with the 1.4.17 release, we could directly open Dicom files from the Conquest data directory with ImageJ, but this no longer works. ImageJ now throws a java.lang.NegativeArraySizeException error when we try to open the files.

However, if we first export the images from Conquest 1.4.19 to a zip file (anonymized or not), they can then be opened without any problems.

The Conquest server is configured to store images in the standard DCM, uncompressed format.

Has there been some change to how Conquest saves images in 1.4.19?

Thank you for your assistance.

Hi Marcel,

Sorry for the delay. I compared the files for an image series stored by Conquest and the Dicom exported to a zip file. The dcm in the zip file is larger by 204 bytes.

Below is the error from ImageJ

ImageJ 1.50i; Java 1.7.0_67 [64-bit]; Windows 8.1 6.3; 94MB of 6000MB (1%)
 
java.lang.NegativeArraySizeException
	at ij.plugin.DicomDecoder.getString(DICOM.java:293)
	at ij.plugin.DicomDecoder.getHeaderInfo(DICOM.java:797)
	at ij.plugin.DicomDecoder.addInfo(DICOM.java:716)
	at ij.plugin.DicomDecoder.getFileInfo(DICOM.java:534)
	at ij.plugin.DICOM.run(DICOM.java:92)
	at ij.IJ.runPlugIn(IJ.java:186)
	at ij.IJ.runPlugIn(IJ.java:169)
	at ij.io.Opener.openImage(Opener.java:319)
	at ij.io.Opener.openImage(Opener.java:243)
	at ij.io.Opener.open(Opener.java:110)
	at ij.io.Opener.openAndAddToRecent(Opener.java:292)
	at ij.plugin.DragAndDrop.openFile(DragAndDrop.java:192)
	at ij.plugin.DragAndDrop.run(DragAndDrop.java:159)
	at java.lang.Thread.run(Thread.java:745)

Incidentally, I tried another viewer, MicroDicom, and was also unable to open the files from the data folder. I'm not sure whether it uses the same libraries as ImageJ. Hope this helps.

Hi Marcel,

An update on this issue.

We downgraded to 1.4.17 to get around this problem. We recently tried the latest 1419a release, and the problem still exists. To answer your previous question, no, we are not storing images in V2 format, just standard uncompressed dicom. Running on Windows 7.

I've made some anonymized examples. I did a quick comparison and the only differences appear to be in the headers (in lines 6, 8, 11, 12, 13,14, 15, 16, 17, 18, 20, 21, 22, and 26). The affected lines were slightly different for the non-anonymized files. The file from the data folder has some missing bytes/characters, as well as some transpositions/substitutions. (UI -> IU, UN -> LO, etc.)

I would like to send the files (~2mb each) to you via pm. I don't currently have such access on the forums.

Thank you for your assistance, and continuing effort with this software.

Hi Marcel,

The update seems to fix the problem we were having. We can now open files directly from the data folder in ImageJ.

Thank you for your help resolving this issue.