Removing burned-in data

[DanielHeath]

I'm looking at adding support for blanking out regions from encapsulated image data. I'd like to upstream as much of that work as possible.

dicom-parser includes the following relevant methods:

• getUncompressedImageFrame
• readEncapsulatedImageFrame
• createJPEGBasicOffsetTable
• readEncapsulatedPixelDataFromFragments

These are used by the wadoURI functionality in the WADO loader.

To support blanking out parts of images, I'd like to extend these methods to tag the source location

I'd like to extend these by tagging the results with the offsets in the underlying byteArray.

Because images can be fragmented, every pixel would need its own offset specified. This would roughly double the memory used to load an image, so it needs to be off by default.

Assuming I can come up with a suitable API / design, would a PR adding this functionality be welcome?

[Neuroforge]

Did you make any progress with this?

[DanielHeath]

Not yet, but I expect I'll need to do it in the next few months.

I've confirmed blanking-out is the right approach since it preserves registration between stacks.