Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Accession Number in Worklist

591 views
Skip to first unread message

or...@volcanomail.com

unread,
Sep 5, 2007, 2:44:59 AM9/5/07
to
Hello all,
I've found a Modality worklist SCP implementation creating more
Worklist entries sharing the same Accession Number. I'm not sure it is
correct.
If I'm wrong and the same Accession Number can be found in many
Worklist entries, which other field univocally identifies a Worklist
entry?
Thank you

Ciro

luigipb

unread,
Sep 5, 2007, 4:54:16 AM9/5/07
to

It is the Scheduled Procedure Step ID (0040,0009). You can have many
Schedule Procedure Steps composing a given Requested Procedure (a
given order), and they obviously share the same Accession number (and
Study Instance UID).

For a better unserstanding of the relationship among the various
identifiers, I suggest to refer to the IHE Radiology Technical
Framework (www.ihe.net).

Regards.

Luigi Pampana-Biancheri
luigi.pampana at esaote.com

or...@volcanomail.com

unread,
Sep 5, 2007, 5:31:39 AM9/5/07
to

Dear Luigi,
thank you for answering.
In the Offis DICOM Basic Worklist Management CTN I've used so far to
test my SCU, it is possible to create many Worklist entries with the
same Scheduled Procedure Step ID, while a unique Accession Number is
automatically assigned to every entry. I think I miss something, I
will check the IHE documentation.
Best regards

Ciro


AnibalJ

unread,
Sep 5, 2007, 6:49:21 AM9/5/07
to

hmmm, it's my understanding that the combination of Accession Number
and Requested Procedure ID and Schedule Procedure Step ID will give
you a unique row in DMWL, and each one will have its own Study
instance UID (they will not share the same stuinsuid, they will only
share the same accession number).

That's what I've seen out in the field and that's what I understood
from IHE's TF.

I wish the TF were much clearer about this situation though.

-Anibal

luigipb

unread,
Sep 5, 2007, 8:51:33 AM9/5/07
to
On Sep 5, 12:49 pm, AnibalJ <anib...@gmail.com> wrote:
>
> hmmm, it's my understanding that the combination of Accession Number
> and Requested Procedure ID and Schedule Procedure Step ID will give
> you a unique row in DMWL, and each one will have its own Study
> instance UID (they will not share the same stuinsuid, they will only
> share the same accession number).

As you can see in the DICOM standard (PS 3.4-2007 Page 189), each item
of the MWL is a Scheduled Procedure Step, so it is quite obvious it
should have a different Sched. Proc. Step ID: no need to consider the
combination with other attributes.

As far as I know (after four Connect-a-thon's), the Sched. Proc. steps
belonging to the same Requested Procedure have the same Study Instance
UID. For example, for the Study ID, IHE suggests to use the Requested
Procedure ID, and in my understanding this means that a Requested
Procedure, made by one or more Sched. Proc. Steps, corresponds to a
study (that is, to an order, that is, a single report will be
produced).

Anyway, it is not a good thing to write your software believing that
the Sched. Proc. Step ID (or any combination of attributes) is unique,
just because I've seen a lot of badly configured worklist server
(especially PACS Brokers) that manage these attributes in a very
incorrect way.

In any case, a modality (that is, the MWL SCU) does not have to care
too much about Accession Number: it is just a matter of managing its
propagation to the MPPS and IODs according to the rules of IHE,
without making too many assumptions.

eric.g...@gmail.com

unread,
Sep 5, 2007, 5:43:39 PM9/5/07
to
There is a lot of history and grief over the Accession Number,
Requested Procedure ID and Schedule Step ID.

Accession Number was originally defined as a "one per study"
identifier in the DICOM image identifying the procedure and/or order
for which the images in the study were acquired. For many exams and in
many institution this one-to-one relationship between the order is
sufficient for their workflow and business processes; but there are a
number of cases where it is not. If you look at the associated
modules in the MWL information model, you'll see the Accession Number
is associated with imaging service request (aka "the order") but an
order may actually include multiple separate requested procedures, and
a procedure itself may include multiple steps which could be scheduled
at different times and be performed on different pieces of equipment/
modalities.

The schedule step is associates a specific acquisition AE title and
date/time for the acquisition. According the IHE guidance, the
scheduled step id should really be nothing more than a sequence number
-- ie step #1, step #2, step #3.

In the one-order=one-procedure model, and in many institutions, the
same value is used for Accession Number and Requested Procedure ID.
However, there are cases and systems which bundle multiple different,
practically unrelated procedures into a single order. The Requested
Procedure is really the desired definitive identifier. It should
correspond to the identifier which will be used for the report
associated with the images. There are cases where a single set of
images may have multiple requested procedure ids. In this case, there
should be multiple reports, or multiple report sections which all
reference back to the same set of images. In a one-to-one
institution, it is fine for the accession number and requested
procedure to be the same. The cases you were seeing where there were
multiple instances of the Accession Number on the MWL either
corresponded to a) procedure with multiple acquisition steps such as
amulti-modality acquisition (some facilities routinely acquire a CR/DX
xray with radfluoro studies). Different AE Title, different Modality
for these steps - the devices may require separate schedule steps in
order to "see" them (many modalities only query for specific modality
codes or for their specific AE Title).

Somewhat stream of conciousness - hope this helps. For further
reading, check out the IHE Radiology Technical Framework. Look at the
group case discussion and mapping of Order Filler HL7 message fields
to the MWL scheduled step attributes


AnibalJ

unread,
Sep 7, 2007, 12:36:12 PM9/7/07
to
On Sep 5, 8:51 am, luigipb <luigi.pamp...@gmail.com> wrote:

...

> As you can see in the DICOM standard (PS 3.4-2007 Page 189), each item
> of the MWL is a Scheduled Procedure Step, so it is quite obvious it
> should have a different Sched. Proc. Step ID: no need to consider the
> combination with other attributes.

...

> Regards.
>
> Luigi Pampana-Biancheri
> luigi.pampana at esaote.com

You're right. My bad. The schedule procedure step has nothing to do
with the study instance UID. I just want to re-iterate one more thing
though to make it clear (and Eric did a good job at making it more
clear as well):

Single imaging acquisition request -> single order (or procedure
really)
Accession Number = A1
Requested Procedure ID = whatever (it really doesn't matter, it could
be A1, or it could be just 1)

Single imaging acquisition request -> many orders (or procedures
really)
Accession Number = A1
Requested Procedure ID = R1

Accession Number = A1
Requested Procedure ID = R2

Accession Number = A1
Requested Procedure ID = R3

etc etc etc

Each requested procedure has its own description, so if you were doing
a CT CHEST/ABDOMEN/PELVIS you could do the following:

A1|R1|CHEST
A1|R2|ABDOMEN
A1|R3|PELVIS

and depending on the MWL SCU, it could be displayed in 3 different
rows, all showing the same accession number (A1), but three different
RPIDs (which most DMWL SCUs don't display), and three different
procedure descriptions.

-Anibal

0 new messages