MPEG2, ?acceptable alternative to DICOM??
R Brad Stamm
digitally as MPEG2 instead of DICOM. This is much more resource efficient
but highly compressed. The manufacturer claim near identical quality
compared to true DICOM when viewing cardiac echos. Anyone have an opinion??
Thanks
--
R Brad Stamm
David Clunie
Hi Brad
To some extent your question is comparing apples and oranges.
MPEG isn't an alternative to DICOM, it may be used to augment DICOM.
DICOM specifies a lot of things, including requirements on the data that
accompanies the pixel data (ie. patient/study/acqusition parameters) as
well as offering uncompressed and compressed (RLE or lossless or lossy
JPEG) encoding.
MPEG2 just addresses compression of the images.
One could encode the DICOM pixel data using MPEG2, and call this a private
transfer syntax and use it between two applications that know about the
private transfer syntax.
Right now, the DICOM ultrasound working group is considering adding a new
standard transfer syntax for MPEG, but not much work has been done yet. A
work item for this effort was approved by the DICOM Committee last year
or the year before, but no one has championed it since.
Now let's separate the question of the form of encoding and encapsulation
of the compressed data stream from the question of DICOM vs. MPEG, from the
quality of the lossy compression scheme itself.
The manufacturer may be claiming that MPEG2 compression compared with so
called "motion JPEG" compression gives superior quality for the same bit
rate (compression ratio). That may or may not be true. There is a body
of literature in the professional journals that addresses this subject
(e.g. from the Cleveland Clinic) that you can consult. I don't have the
references to hand, but a medline search would find them.
Alternatively, the manufacturer may be claiming that MPEG2 compression
gives satisfactory results compared to lossless compression or no
compression at all. That subject is also addressed in the literature.
Regardless, it isn't a DICOM issue ... it is a compression scheme choice
issue. Any compression scheme (almost) could be encapsulated in DICOM to
gain the benefit of the accompanying management and annotation information.
Your manufacturer may well be right if their claim was modified to read
"... near identical quality compared to uncompressed images when viewing
cardiac echos" without mentioning DICOM. Of course choice of compression
ratio makes a difference ! What I presume they mean by "true DICOM" is
either "compared to that subset of DICOM that our competitors may have
implemented in their products" or "those forms of lossless or lossy
compression currently defined as standard transfer syntaxes in the
DICOM standard".
If you just spit out an MPEG stream without any additional information
though, you loose all the benefits of DICOM ... ie. accompanying stuff
that helps you find, manage, track and display images.
Anyone who feels MPEG should be added to DICOM as a standard transfer
syntax should come up with a technical proposal and join the working
group. If MPEG is important to the echo community, then the constraints
on it and manner of its encapsulation should be specified, to promote
interoperability.
To reiterate, DICOM neither restricts one to uncompressed or lossless
compressed images, nor to any particular lossy scheme, nor does it
recommend any particular scheme for any particular purpose. It does
specify a set of alternatives that can be expanded as the need arises.
david
--
David A. Clunie mailto:dcl...@idt.net
Director, Medical Imaging Technologies http://idt.net/~dclunie/
Quintiles Intelligent Imaging http://www.i2image.com/
521 Plymouth Rd #115 Work 610-238-0572 Fax -0578
Plymouth Meeting PA 19462 Home 717-897-7123 Fax -5117