Request: Add support for OCTA data in .e2e files

[23dhruvs]

Hi,
First of all, thank you for making and maintaining this library, and for the recent addition of support for OCTA data from .vol files — it has been very helpful!

I wanted to ask if similar support can be added for .e2e files. Based on my exploration of the file structure, OCTA data in .e2e files is stored in chunk type 1073751901, which is different from the chunk used for standard OCT + fundus image data.

I tried adapting the existing logic by pointing the parser to this OCTA chunk (without additional transformations). Using the oct-converter library for .e2e → DICOM conversion, I was able to extract OCTA images, but they consistently come out at half the width of the corresponding OCT scans.

Here is a sample b-scan and corresponding angiography data extracted with dimensions from header file and after halving the width.

b-scan	angio	half width angio

My questions:

1. Have you encountered this issue with OCTA data from .e2e files (or from other Heidelberg file formats)?
2. Is the halved width an artifact of the data structure itself, or might it require a different decoding/transformation step compared to OCT?
3. Would it be possible to extend eyepy to natively read OCTA data from .e2e files, similar to how .vol support was recently added?

Thanks again for maintaining this library!

[Oli4]

Hi,

thank you for reaching out. I currently do not plan to extend the E2E support and would recommend using the VOL export instead. The E2E data probably holds the raw data and requires additional processing. I am happy to help with a pull request in case you want to take on this challenge.

With regard to the width issue, maybe the data is encoded in a different format float16 instead of float32 for example.

Best Olivier

[23dhruvs]

Hi, thank you for your swift response - your suggestion solved the width issue.

Unfortunately, we can’t use VOL export, as the Spectralis devices we’re working with are older and don’t support that option. Since we’ll be working with E2E files only, we’ll focus on adapting support for OCTA from E2E and will try to contribute our progress back to the eyepy project.

Thanks again for your guidance!