US region Calibration
b-rama...@mustafasultan.com
I have problem in interfacing Ultrasound to PACS interms of meaurment.
The Ultrasound sends the image with US Regional Calibration module and
the PACS vendor doesn't support this module. The PACS vendor request
pixel value with image.
Because of this when we send the image to PCS and the customer measure
the image in PACS workstation they get teh value in pixels and not in
cm.
The Ultrasound vendor says teh US region Calibration should be used for
Ultrasound image as per NEMA spex in Ultrasound images and not pixel
values.
But the PACS vendor says both pixel and US region calibration can be
used for ultrasound image and US region calibration technicque may
produce worng values and better to use only Pixel values for image
transfer for Ultrasounf too.
Please advise which statment is corrrect and is there any document from
NEMA site I can refer for the confirmation. Why we should use US region
calibration technique for US images only and what is the siginificant.
Also if teh PACS vendor doesn't have teh facility to accept teh US
egion calibration then is there any work around to measure the image in
PACS in cm.
Kindly help.
Thanks and regards,
Ram
eric.g...@gmail.com
If the PACS claims to support ultrasound, they really should support
the US Region Calibration module. This is one of the grey areas of
compliance that really doesn't surface very well in a DICOM conformance
statement. PACS applications will claim they "support" the SOP class
if they can receive it via DICOM and can display the image pixel
values. Some vendors have differentiated themselves by explicitly
calling out the fact that they support region calibration - but nothing
in the conformance forces them to declare their support (or lack
thereof) to support the spatial measurements.
Although it isn't as accurate as you would get with the PACS using the
calibration information from the vendor, many ultrasound vendors also
include a distance calibration marker in the image pixel data --ie a
distance scale along the side boundaries of the triangular transducer
echo return. Many PACS applications provide an "image calibration"
operation where a user can mark out a line on the image and enter its
length value in mm or cm. The PACS application will then apply that
distance scaling across all the image pixels so any additional
measurements in the image use the same pixel to distance conversion
scaling.
So, a user viewing an ultrasound image with one of these distance
scales in the image can mark out a distance on the provided scale and
enter whatever distance they have marked out. Most radiologists
wouldn't be pleased to do this, but might have one of their technicians
do this operation on the exam before the radiologist reads
b-rama...@mustafasultan.com
Thanks for the reply.
But as I understand with US region calibration images from Ultrasound
needs to be calibrated with teh scale which the US people send for each
image , is this correct.
OR
If teh PACS vendor do the calibration in one image converting the
pixels to cm then it will be applied uniformly to all the images
receivd fromt hat modality.
Further do any one have the soft copy or the url for the following
dicom stament which describes the US region calibration,i.e,
Part 3 of the DICOM Standard, Table A.1-1, and section A.6.4. DICOM
Tag 0028,0030 Pixel Spacing is only used in the Image Plane and NM
Image Pixel modules, which are not used in Ultrasound.
Thanks in advanse.
reg,
Ram
eric.g...@gmail.com
calibrated using the digital logic designed into the device by the
manufacture. PACS/display applications that read these will be able to
automatically translate distances or time or doppler values encoded
into the pixels into their digital eqivalents without any manual
calibration functions by the user.
The manual calibration by a PACS using the reference scale displayed
into the image will only be valid for the image in which the user
performs the calibration. It won't (shouldn't) apply to any other
images in the same series, study or even greater scope. This seriously
limits its usefulness
Thomas Freier
for such cases we have developed the DAC (DICOM Attribute Converter).
It can grab the message from the network and change it before sending
to the PACS... and vice versa. So depending on what your PACS is able
to store we might be able to move the Region Calibration Module to
somewhere else (e.g. private Tags) when sending to PACS. When
retreiving from PACS we get that attributes and copy them back.
Have a look at the link below:
http://www.soft-gate.de/EV/index.html
To be sure we can solve it using this tool I would need the conformance
statements of both applications.
Thomas
b-rama...@mustafasultan.com
This means most of the PACs vendors doesn't support US region
calibration in their PACS for ultrasound image thats why we are getting
the measurment in pixels and not in cm .
To overcome this short fall they ask the radiologist to do the
calibration with the standard length which all the radiologist not
interested in the same.
Could you advise which PACS supports from various well known PCAS
vendors, like, Philips, GE, Siemens, AGFA, KODAK etc , the US region
calibration and working perfectly.
Is anyone could tell us why we should use US region calibration only in
Ultrasound Image, what is the siginificance of this, I couldn't find
this in teh NEMA documentation, as I am not the expert in reading theis
DICOM staments.
If any one has this could you share this with me.
Further one of my collegue asked to check the Part 3 of the NEMA
Standard, http://medical.nema.org/dicom/2004/04_03PU3.PDF , for teh
table Table A.1-1
and section A.6.4. ,
when I checked the table, it shows IDOs module against Different
Modalities. For US region calibration IOD it shows against ultrasound
only as U. But what U stands for in this table couldn't find the
definition.
Also in this table they use other letters like, M & C, but couldn't
find the expansion of the same.
If you know could you let me know what this letters indicate.
Further in NEMA documentation is there any place they mention cleraly
that we should use only US region calibration for Ultrasound image, if
yes, do send me that link to enable me to take a printout.
Thanks in advance.
Ram
lui...@hotmail.com
an informative table, see the various detailed tables for the various
IODs), U means "optional" module (for the device that creates the SOP
Instance), M means "mandatory" and C means "conditional" (that is, if
the condition is satisfied the module must be present, otherwise it
must be absent, unless explicityl allowed that the module can be
present even when the condition is not meet). Do not ask me where I
found this explanation of U/M/C, I am quite confident somewere in the
standard the explanation is present :-)
Normally, very popular images (CT, MR, etc.), use the (0028,0030) Pixel
Spacing DICOM attribute (included in the Image Plane Module, see Table
C.7-10 of the PS3.3 - 2004 book of the DICOM Standard) to carry the
information about the spatial calibration of the pixels, that is the
horizontal and vertical distance between the pixels of the image. This
is because these images can only be spatial planes. The Image Plane
Module is enclosed in most DICOM IODs (Information Object Definitions,
that means the definition of the format of the DICOM images), like CT
and MR, as I wrote. BUT the Image Plane Module it is NOT included in
the Ultrasound and Ultrasound Multiframe images (see Table A.6-1 of
PS3.3), so you will NOT find it in any US or US-MF image, so you will
NOT find the (0028,0030) Pixel Spacing attribute.
This is because the US and US-MF images normally have a more complicate
planar structure than a CT or MR: for every US image you can have
different regions with different calibrations. For this reason,
instead, in the US you will find a more complicate calibration scheme,
the Ultrasound Region Calibration module. Using the information
contained in this module it is perfectly possible to make measurements
over the various regions of the image without having to manually
calibrate any of them. Not only planar distances over the regions that
represent a spatial plane.
Any conversion tool is useless (as I said, not all the pieces of an US
image are spatial regions !), just study how the US Region Calibration
module works, and implement it. Otherwise the customer of the PACS will
have to manually decide the position of the spatial regions, and to
calibrate them (provided a ruler is burned in the pixels, but this is
not mandatory for the modality !), and she will not be able to perform
any measure on the regions that are not spatial ones (that is, where
rows and columns do not both represent spatial axes).
Regards.
Luigi Pampana-Biancheri
=================================================
Luigi Pampana-Biancheri, BS in Physics
ESAOTE S.p.A.
DICOM and Connectivity Management (EDMG)
via Siffredi 58 - I-16153 Genova (Italy)
mailto:luigi....@REMOVETHISesaote.com
http://www.esaote.com
=================================================
b-rama...@mustafasultan.com
Thanks for the detail explanation.
But the reputed supplier like Siemens in their PACS not supporting US
Region Calibration and because of this the meaurment at PACS comes in
pixels. They say Ultrasound should send the image with pixel values to
display correctly the meaurment in cm.
Furthe the Siemens ultrasound sends the image with pixel spacing values
and it's working OK in their PACS and this happens with Toshiba
ultrasoud too. Since these supplier of Ultrasound they provide
workaround in their DICOM to solve the problem of meaurment in PACS
where the PACS vendor doesn't have the US region calibration module.
If NEMA makes it as part of DICOM 3.0 standard that only US regin
calibration should be used for US images then this confusion at sites
will change and I hope the PACS vendrs like Siemens and others wil
implement this technique soon in their PACS.
Could you name the PACS vendors int eh industry who have this technique
of US Region Calibration , other than GE,
reg,
Ram
lui...@hotmail.com
> Region Calibration and because of this the meaurment at PACS comes in
> pixels. They say Ultrasound should send the image with pixel values to
> display correctly the meaurment in cm.
This is just a trick, another way to create a DICOM dialect. As I
wrote, using the pixel spacing is not the correct way to carry this
information.
> If NEMA makes it as part of DICOM 3.0 standard that only US regin
> calibration should be used for US images then this confusion at sites
> will change and I hope the PACS vendrs like Siemens and others wil
> implement this technique soon in their PACS.
This is what DICOM already says: pixel spacing is not inclueded in the
US and US-MF IODs. You can add it making it an extended SOP Class, as
you can add any other standard or proprietary attribute, but this is
not intended for inter-operability, of course.
> Could you name the PACS vendors int eh industry who have this technique
> of US Region Calibration , other than GE,
I cannot do a complete survey, anyway I personally verified it with
Kodak DirectView 5.2 and Agfa (adding a sw patch). Also the Esaote
ultrasound miniPACS and viewer (BioPACS) obviously supports it. And
probably many more, especially the ultrasound oriented PACSs and
viewers.
Regards.
Luigi P.B.
David Clunie
Here is what I wrote in response to someone else asking
about this earlier today:
"The Pixel Spacing attribute is defined in the CT and MR IODs,
for example, where it has a specific meaning.
It is not defined in the US IODs, and therefore when added
by an acquisition device as a Standard Extended SOP Class,
there is no standard meaning for its use.
This has been a huge problem for CR images, for example,
which have a similar problem ... vendors have added the
attribute and implied something about its meaning, but
even more meaningless than with US because the spacing
in the patient is impossible to determine with projection
radiography, unlike cross-sectional modalities, because
of the diverging x-ray beam and the depth of any location
of interest within the patient. Hence CP 586.
Workstations that blindly ignore the IOD (SOP Class) and
make expedient use of Pixel Spacing (or Imager Pixel Spacing
or any other spacing attribute) without respect to the IOD-
specific meaning are in my opinion creating a safety
hazard, because they have no idea how the modality vendor
might have encoded the attribute and for what purpose."
I am considering proposing a CP that FORBIDS the addition of
Pixel Spacing attributes as a Standard Extended SOP Class
to inappropriate IODs like US that already have a spacing
mechanism defined, which lame and useless PACS workstation
vendors should be using.
It constantly mystifies me why customers tolerate PACS
that do not implement modality-specific functionality
on the display side such as is necessary for ultrasound
or nuclear medicine.
No matter how "reputable" the supplier is in a general
sense, as Ram used the term, that does not mean that
their PACS is any good for "niche" (sic) applications like
Ultrasound, which have "special" requirements.
As an implementor, I hate US Region Calibration, and ultrasound
weirdness in general, as much as the next person, but that does
not mean I would purchase or recommend a PACS workstation that
didn't support it (properly).
David
lui...@hotmail.com wrote:
>>But the reputed supplier like Siemens in their PACS not supporting US
>>Region Calibration and because of this the meaurment at PACS comes in
>>pixels. They say Ultrasound should send the image with pixel values to
>>display correctly the meaurment in cm.
>
> This is just a trick, another way to create a DICOM dialect. As I
> wrote, using the pixel spacing is not the correct way to carry this
> information.
>
>>If NEMA makes it as part of DICOM 3.0 standard that only US regin
>>calibration should be used for US images then this confusion at sites
>>will change and I hope the PACS vendors like Siemens and others wil
b-rama...@mustafasultan.com
Could you let me know whether you can addres this problem with your
available solutions and if yes what will be the cost.
reg,
ram