unable to view hologic tomosynthesis images

[Rajul Shah]

Having issue with viewing images using either oviyam or weasis (latest versions) with tomo files.  These are compressed using J2K loseless. using fedora 21 with chee2.18.3 jre 1.7.0_79-b15 and latest jai libraries (i.e., jai_imageio-1.2-pre-dr-b04.jar).  Have more than enough memory/cores (32g/8+core).  The compression service is serialized with 4g allocated.  My jvm setting are:

JAVA_OPTS="-D64 -Xms8g -Xmx12g -XX:MaxPermSize=256m -XX:+UseParallelGC -Dsun.rmi.dgc.client.gcInterval=3600000 -Dsun.rmi.dgc.server.gcInterval=3600000"

link to anonymized file: https://dl.dropboxusercontent.com/u/84157163/patient.zip

consistently crashes jvm after first 2-3 display of images.  Any pointers to debugging this issue please.  We have tried this on various other hardware and OS configuration (ubuntu) with same results.

On TRACE, here is error message: 

*** Error in `/usr/java/jdk1.7.0_79/jre/bin/java': double free or corruption (out): 0x00007fca1c49c460 ***
======= Backtrace: =========
/lib64/libc.so.6(+0x7d19d)[0x7fca5d13819d]
/usr/local/chee/bin/native/libclib_jiio.so(+0x31097)[0x7fca3f128097]
======= Memory map: ========
00400000-00401000 r-xp 00000000 08:01 52650715                           /usr/java/jdk1.7.0_79/jre/bin/java
00600000-00601000 rw-p 00000000 08:01 52650715                           /usr/java/jdk1.7.0_79/jre/bin/java
00c43000-00cbf000 rw-p 00000000 00:00 0                                  [heap]
770000000-775000000 rw-p 00000000 00:00 0 
775000000-780000000 rw-p 00000000 00:00 0 
780000000-7aab00000 rw-p 00000000 00:00 0 
7aab00000-7d5500000 rw-p 00000000 00:00 0 
7d5500000-800000000 rw-p 00000000 00:00 0 
7fca04000000-7fca04023000 rw-p 00000000 00:00 0 
7fca04023000-7fca08000000 ---p 00000000 00:00 0 
7fca0bafb000-7fca0bafe000 ---p 00000000 00:00 0 
7fca0bafe000-7fca0bbfc000 rw-p 00000000 00:00 0                          [stack:5680]
7fca0bbfc000-7fca0bbff000 ---p 00000000 00:00 0 
7fca0bbff000-7fca0bcfd000 rw-p 00000000 00:00 0                          [stack:5679]
7fca0bcfd000-7fca0bd00000 ---p 00000000 00:00 0 
7fca0bd00000-7fca0bdfe000 rw-p 00000000 00:00 0                          [stack:5678]
7fca0bdfe000-7fca0be01000 ---p 00000000 00:00 0 
7fca0be01000-7fca0beff000 rw-p 00000000 00:00 0                          [stack:5677]
7fca0beff000-7fca0bf02000 ---p 00000000 00:00 0 
7fca0bf02000-7fca0c000000 rw-p 00000000 00:00 0                          [stack:5676]
7fca0c000000-7fca0c256000 rw-p 00000000 00:00 0 
7fca0c256000-7fca10000000 ---p 00000000 00:00 0 
7fca10000000-7fca10154000 rw-p 00000000 00:00 0 
7fca10154000-7fca14000000 ---p 00000000 00:00 0 
7fca14000000-7fca14288000 rw-p 00000000 00:00 0 
7fca14288000-7fca18000000 ---p 00000000 00:00 0 
7fca18000000-7fca19b2c000 rw-p 00000000 00:00 0 
7fca19b2c000-7fca1c000000 ---p 00000000 00:00 0 
7fca1c000000-7fca20000000 rw-p 00000000 00:00 0 

[Matthew Hill]

http://stackoverflow.com/questions/22491797/java-double-free-or-corruption  ??

Error in `/usr/java/jdk1.7.0_79/jre/bin/java': double free or corruption (out):

--
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.

[Dr Suresh Viswanathan]

It would help if you could share an anonymised DICOM image for testing. 

Suresh. 

Sent from my iPhone

--

[Nicolas Roduit]

It works without any problem directly in Weasis.

Are you using the right native library in dcm4chee ? see this post.

I've also updated the installation wiki page.

[Rajul Shah]

I think we have it figured out.  The Hologic was setup to send JPEG2000 loseless; it might be using different library to compress.  With sending as uncompressed turned on, and letting chee do compression/decompression, no issues so far with Oviyam or Weasis view

Thanks 

[Nicolas Roduit]

If it is Hologic, there are different output formats: CTO, BTO et SCO. Only the last one seems to be compliant with the DICOM standard (the two first can be considered as proprietary formats).