Multi-Frames with ES

[greggy]

It would appear from the standard that there is no concept of a multi-
frame ES image IOD.

However, I can attach pixels of a single buffer of the size to cover 2
frames, set the NumberOfFrames (28,8) to 2, and all is fine - except
that only my own viewer will acknowledge the additional frame (that I
know of so far). The IOD transfers back and forth to and from the PACS
just fine. Both frames are present in the object.

Is stating multiple frames for ES illegal or just not specifically
defined?

[David Clunie]

It it is not a question of the Modality value, but rather the
IOD (and SOP Class). The latter may constrain the former,
not the other way around.

The VL Endoscopic Image IOD is a single frame image IOD.

I.e., it is illegal in this object to have multiple frames and
illegal to have Number of Frames present at all, and certainly
not if it has a value other than 1.

Whether it works with some PACS or viewers is irrelevant; it is
illegal, and hence likely not to work with other systems.

Any IOD that allows for multiple frames will contain a module
that includes Number of Frames as an attribute, e.g., the
Multi-frame Module.

You could consider using the Video Endoscopic Image IOD instead
of the VL Endoscopic Image IOD, but you may encounter problems
because some systems may assume that the frames will be MPEG
encoded, though there is no such restriction in the standard.

As I have mentioned before, since you are not doing endoscopy
in the first place, I think you might be better off using the
Multi-frame True Color Secondary Capture Image IOD instead,
which is multi-frame, does not constrain the value of Modality.

Also, if you are only going to send two frames, you would
probably have greater interoperability if you sent these as
separate single frame objects, e.g., as two ordinary Secondary
Capture Image objects.

From what requirement does the need for multiple frames arise ?

Are these video-like, multiple spatial locations, etc. ?

David

PS. Have you tried running dciodvfy or some other validator on
the instance you are creating ? Your goal should be to eliminate
all errors and warnings from such validators, and dciodvfy
should detect when the number of frames is incorrect (let me
know if it doesn't).

[greggy]

Our device takes a picture and does so using two different types of
lighting. Our device (and PC software) require both frames always. It
has been problematic for us to manage dual frames in separate objects,
due to the "what if.." scenarios that we imagine "will" occur in the
field. For us, it is far simpler to just put them both in a single
object.

Having said that, I have implemented the SC MF TC and it works fine. I
am finishing the clean up work now.

I will look into the tool you suggest to verify my dataset.

Thanks a lot for your time on this.