Re: Is compressed pixel data in unencapsulated format valid?

On Nov 12, 2:30 pm, Michael Doppler <m.dopp...@xxxxxxxxx> wrote:
We have received a sample image from a customer that is compressed
using JPEG lossless non hierarchical 1st order prediction transfer
syntax and includes the pixel data in unencapsulated format, without
encoding it in a fragment. Is this valid according to the DICOM
Standard?

In chapter 8.2 of DICOM Standard 2008, Part 5 [1] it sounds as if
compressed pixel data shall be encoded in at least one fragment in
encapsulated format and not present in unencapsulated format. Is this
correct?

Thanks in advance,
Michael

[1]ftp://medical.nema.org/medical/dicom/2008/08_05pu.pdf#page=47

8.2 NATIVE OR ENCAPSULATED FORMAT ENCODING

Pixel data conveyed in the Pixel Data Element (7FE0,0010) may be sent
either in
a Native (uncompressed) Format or in an Encapsulated Format (e.g.
compressed)
defined outside the DICOM standard.

...

The encapsulated pixel stream of encoded pixel data is segmented in
one or
more Fragments which convey their explicit length.

No you need at least two items. The first one is the Basic Offset
Table (may be empty), the second one is indeed the jpeg stream. Anyway
simply storing the JPEG stream within Pixel Data element (7fe0,0010)
is a *very* bad idea. This is really a bad implementation of the DICOM
standard, just run away quick !

One way to check is run gdcmconv --raw or dcmtk: dcmconv +ti on the
file.

As a side note, I have seen people doing this kind of things, but they
were storing the JPG in a Raw Data Storage.

HTH
-Mathieu
Ref:
http://gdcm.sourceforge.net/html/gdcmconv.html
.

Relevant Pages

  • Re: pixel data chunks
    ... all of the pixel data is put into a single block in the file. ... The rationale for defining the format this way was probably that it ... permits encoders to run with a fixed-size memory and still ... The only encoding where you can safely expect this encoding ...
    (comp.protocols.dicom)
  • Re: PGM file
    ... >>in the header of PGM file ... > One indicates that the pixel data is encoded as (portable but big and ... > slow) ASCII values, while the other indicates raw byte data that is ... There is no endian issue with the 8-bit format. ...
    (sci.image.processing)
  • Re: pixel data chunks
    ... The pixel data in these images are divided into chunks of<= 65532 ... The rationale for defining the format this way was probably that it ... Whenever the memory buffer is ... encapsulated transfer syntaxes. ...
    (comp.protocols.dicom)
  • Re: loading bitmaps from files
    ... > I'm writing a programme which needs to load bitmap files from ... > disk and then access the raw pixel data. ... > that raw data as one plane, ... If you want 32bpp format, you have to ask for it. ...
    (microsoft.public.vc.language)
  • Re: Urgenr! JPEG or Default Transfer Syntax to Choose?
    ... .I am getting images scanned from a jpeg scanner and s ... does not encodes the actual pixel data though it can ... Add all the tags as per selected IOD to DICOM ... pixel that as per jpeg transfer syntax though it is jpeg ...
    (comp.protocols.dicom)