Error in IHE Technical Framework Vol III

Stephan

unread,

Sep 24, 2008, 4:04:04 AM9/24/08

to

Hello,

I found an error in IHE Technical Framework Vol. III but I don't know
where to report. Can someone please tell me where should I address
this?
Thanks
Stephan

David Clunie

unread,

Sep 24, 2008, 6:36:21 AM9/24/08

to

Send it to me.

David

Q8PACS-ADMIN

unread,

Sep 25, 2008, 7:05:40 PM9/25/08

to

I found an Error in DICOM .. can I contact you , David ?

David Clunie

unread,

Sep 26, 2008, 8:09:54 AM9/26/08

to

Of course.

Hussein

unread,

Sep 30, 2008, 8:59:23 AM9/30/08

to

Hi David,

I noticed one day that some Modalities have different definition for
DICOM tag (0018,0088) "Spacing Between Slices". In fact, modality
measures space between slices as space between internal edges between
two consequent ones. however, DICOM part3 2007 measures the spacing
between the slices as distance between center of two consequent
slices.

In different words, the relation between the two definition is
SS(DICOM) = SS(Modality) + Slice Thickness

Therefore, we will end up exchanging wrong information after words.

Is there any other DICOM tag where SS(MODALITY) is sent ?

If not, Can some body bring this issue to DICOM meetings ?

Thanks David,

Hussein

unread,

Sep 30, 2008, 8:59:36 AM9/30/08

to

Hi David,

I noticed one day that some Modalities have different definition for
DICOM tag (0018,0088) "Spacing Between Slices". In fact, modality
measures space between slices as space between internal edges between
two consequent ones. however, DICOM part3 2007 measures the spacing
between the slices as distance between center of two consequent
slices.

In different words, the relation between the two definition is
SS(DICOM) = SS(Modality) + Slice Thickness

Therefore, we will end up exchanging wrong information after words.

Is there any other DICOM tag where SS(MODALITY) is sent ?

If not, Can some body bring this issue to DICOM meetings ?

Thanks David,

On Sep 26, 3:09 pm, David Clunie <dclu...@dclunie.com> wrote:

David Clunie

unread,

Sep 30, 2008, 12:38:04 PM9/30/08

to

Hi Hussein

There is no "error in DICOM" in this respect ... the meaning of
(0018,0088) is clear (between centers not adjacent edges).

If some modalities are sending a value that is different in that
attribute, then there is an "error in the modality", i.e., a bug,
and you should complain to the vendor.

If you want to compute the spacing between the edges, subtract
Slick Thickness (0018,0050) from Spacing Between Slices (0018,0088).

DICOM will not add a new attribute for "spacing between edges"
since there is no need for it (i.e., it can be computed).

David

Hussein

unread,

Sep 30, 2008, 6:30:55 PM9/30/08

to

David,

I'm saying this is wrong because DICOM definition is not consistant
with practice. I've asked a couple of radiologist about their
definition of Slice spacing and they didn't agree with DICOM
definition Plus DICOM definition is not consistent with the TAG
name .. "spacing between slices"..

If you ask any person about the space between any two things then he
will not agree with DICOM definition..

Thanks,
Hussein

> >>>>> Stephan- Hide quoted text -
>
> - Show quoted text -

David Clunie

unread,

Oct 1, 2008, 4:13:15 PM10/1/08

to

Hi Hussein

The name of the tag is irrelevant; the definition in the standard
is clear and unambiguous and is not going to be changed.

What the user needs/wants to see can be computed from the parameters
provided.

There is only one "practice" here as far as modality engineers are
concerned, and that is to a) encode attributes as DICOM specifies,
and b) convert those appropriate into what the user needs to
see when displayed.

David

Hussein

unread,

Oct 4, 2008, 6:43:57 PM10/4/08

to

Hi David,

I understand you point of view, but so far All radiologists(user) I
met they agree with the modality definition.

Thanks,

David Clunie

unread,

Oct 9, 2008, 6:18:46 PM10/9/08

to

Well, I'm a radiologist, and I don't know of any other radiologist
in our group or specialty (cancer clinical trials), who think in
any other terms than 'slice thickness' and "reconstruction interval',
the latter being the same as the spacing between the centers of
slices. Occasionally, I used to hear folks refer to the 'skip',
the distance between the edges, but not recently.

But like I said, it doesn't matter as long as implementers read
the standard and follow it exactly, and don't encode the 'skip'
in the attribute that is defined to be the distance between
centers.

Robert Horn

unread,

Oct 10, 2008, 11:33:36 AM10/10/08

to

On Oct 9, 6:18 pm, David Clunie <dclu...@dclunie.com> wrote:
> Well, I'm a radiologist, and I don't know of any other radiologist
> in our group or specialty (cancer clinical trials), who think in
> any other terms than 'slice thickness' and "reconstruction interval',
> the latter being the same as the spacing between the centers of
> slices. Occasionally, I used to hear folks refer to the 'skip',
> the distance between the edges, but not recently.
>

Hussein and David,

This is just another example of the ambiguity of English. You can get
different responses by changing how you ask the question. It's not
just in radiology. In carpentry you will say that the studs in a wall
are 24" apart, and then need to remind the newcomers that you measure
center to center not the gap between edges. The same applies for
posts for fences. It is the same issue as radiology slices and the
same ambiguities arise. As long as people read the definition of the
DICOM attribute this ambiguity is removed.

Hussein

unread,

Oct 24, 2008, 9:59:17 AM10/24/08

to

Hi Robert,

This particular problem is not only because of mis understanding. It's
is because of two different definitions. Modality uses different
definition on when you look at the Slice spacing on console, but when
it send through dicom interface it follows the other definition...
this is Fatal !!! Modality Am taking about here is a GE Signa MR, and
GE AW..

Thanks,