Region of interest in vector images

Return to SDTC page; Return to CDA R3 Formal Proposals page.

See CDA R3 Formal Proposals for instructions on using this form. Failure to adhere to these instructions may result in delays. Editing of formal proposals is restricted to the submitter and SDTC co-chairs. Other changes will be undone. Comments can be captured in the associated discussion page.

(This is a template. Do not edit! Copy and paste source into to a new page)

(An announcement of this proposal must be submitted to the Structured Documents list to be formally submitted.)

Submitted by: <<Name>>	Revision date: <<Revision Date>>
Submitted date: <<Submit Date>>	Change request ID: <<Change Request ID>>

Issue

The CDA R2 Normative Web Edition, section 4.3.6.7 states the following:

The units of the coordinate values in RegionOfInterest.value are in pixels, expressed as a list of integers. The origin is in the upper left hand corner, with positive X values going to the right and positive Y values going down.

Using pixels works well for bitmaps, JPEG, GIF and other similar images. However, a vector based image does not use pixels and they can be scaled up an down. Pixels would not work in this case.

Recommendation

There are two options to consider. One of the two could be used, or both could be written into the specification.

1. Specify an image size, then use pixels:

Use something similar to the "regionOfInterest" that is used now. However, for vector objects, require that a set of image dimensions be assigned in terms of pixels. This would define the lower right side of the image and would set the scale that would be used for the overlay. For example, if you want your region of interest overlay to be used for an image the size of 2000 pixels by 3000 pixels, you would define the image size first:

              
              <value value="2000"/>
              <value value="3000"/>

Then you would do your normal pixel references for the overlay as they are currently defined in the CDA R2 specifications. In the example above, "IMAGE_SCALE" is just a place holder code and a better one may need to be defined.

2. Use percentages for the overlay coordinates:

This could be done exactly the same as the current method and just change the code value. For example, if you wanted to define an ellipse, just use percentages for the X,Y pairs. The distance would be relative to the upper left corner of the image still, but it could grow and shrink depending on how large you decide to stretch the image. The only thing you would need to change in this case is the code. Everything after that would be in terms of percentages instead of pixels:

                *** Note:  x,y coordinates now in percentages
              <value value="30"/>  *** Note: x is 30% from the left side
              <value value="10"/>  *** Note: y is 10% from the top
              <value value="30"/>
              <value value="70"/>
              <value value="20"/>
              <value value="40"/>
              <value value="40"/>
              <value value="40"/>

Discussion

Should we also consider 3 dimensional images and/or videos in this proposal?

3D Images

If we do 3D images, we would need to include an extra axis in the overlay and/or have a 2D overlay and include a specific orientation of the 3D image (x,y,z rotation).

Videos

For videos, you may need to include a referenced time frame that the overlay is effective for the particular video. For example, a 2D region of interest overlay is referenced during the video time of 3 seconds to 4.5 seconds.

This all could get very messy, but as 3D imaging and video technology advance, it would be nice to have a standard that can use them.

Resolution

(Resolution is to be recorded here and in the referenced minutes, which are the authoritative source of resolution).