| T O P I C R E V I E W |
| tonev |
Posted - Feb 24 2004 : 2:43:43 PM
Hi There--I've just received some dcm images from Stanford and I'm having difficulty creating bxh files for these volumes. Whenever I attempt to use the bxhabsorb function and select the first image in the series, I am presented with the error message, "Current image is not DICOM_Volume format!". Readmr then defaults to select a raw 64x64x__x1 format. I should also note that when I try to read in the file as a dicom volume using readmrold, it works fine. Any help would be appreciated.
Cheers,
Simon |
| 5 L A T E S T R E P L I E S (Newest First) |
| syam.gadde |
Posted - Feb 25 2004 : 10:45:39 AM
The DICOM file appears to be generated by eFilm, a commercial, proprietary software product that reads/writes/transfers DICOM data. One of the fields (SourceApplicationEntityTitle) has a length of 18 bytes, and the type (AE) of that particular field is specified to have a length no more than 16 bytes. Matlab's DICOM validation fails on that. eFilm may be the culprit, or it could possibly be some other program in the pipeline. I don't know if there is a solution besides having your colleagues at Stanford try other software or contacting the company with their concern. As far as alternative tools, I have heard good things about DicomWorks, though I have not tried it yet. |
| tonev |
Posted - Feb 25 2004 : 10:36:00 AM
Syam--The DOS version did the trick. I must have used the wrong syntax the first time I tried it. Based on your analysis, is there something specific about the imposter DICOM file that I can pass on to Stanford (i.e., something about the parameters they set when creating them)? I'd like to try and get this fixed on their end rather than accomodate it here.
Cheers,
Simon |
| syam.gadde |
Posted - Feb 25 2004 : 10:20:49 AM
Simon,
I tried out your file, and apparently it's not valid DICOM, though it tries very hard to be. Because we use the standard Matlab DICOM parsing functions, we can't change how they behave, so I might suggest using the DOS command-line version of bxhabsorb.
I'm not sure how you got your error, but if I run:
P:\bxh\bin\bxhabsorb.exe *.dcm tmp.bxh
in the same directory as the image file you sent me, it creates a BXH file without complaint. If you still get the error, could you let me know exactly what arguments you're passing to bxhabsorb.exe? |
| tonev |
Posted - Feb 24 2004 : 4:56:53 PM
Thanks Syam--The error dialog now says the following:
Error reading DICOM_Volume format file:
Error using ==> readmr
Error using ==> c:/matlab6p5/toolbox/images/images/private/dicom_read_attr
(read_elt)
Attribute (0002,0016) has inconsistent length 18.
When I try the DOS version, I get the following error:
bxh_DICOM_readFiles: error opening file ('*\.') |
| syam.gadde |
Posted - Feb 24 2004 : 3:38:40 PM
Simon,
That error message you got is misleading, and the underlying error could be almost anything. I have put an updated readmrgui.m in the beta directory (\\gall\programs\User_Scripts\beta). If you add that to your Matlab path and try again, it may give some more info. Please post your results and we'll try to figure out what's going on...
If you get the same error, consider trying the DOS command-line bxhabsorb to see if it also produces an error. Assuming \\gall\programs is mapped to P:
P:\BXH\bin\bxhabsorb.exe --search-for-others myDicomFile myOutput.bxh
|
|
|
