Hello,
I’m encountering an issue when converting an SVS image to DICOM using OrthancWSIDicomizer. After the conversion, the pixel dimensions of the resulting image do not match those of the original file.
Observed Issue
Dicomizer version is linux-standard-base/orthanc-wsi/3.0.
When checking the image before and after conversion with QPath, I noticed a difference in dimensions:
- Original image dimensions (proprietary format): Pixel width: 0.499 µm / Pixel height: 0.499 µm
- Image dimensions after conversion: Pixel width: 0.3261 µm / Pixel height: 0.4557 µm
The original image is an SVS file, compatible with OpenSlide (CMU-1.svs).
Error Message Before Conversion
Before the conversion starts, I get the following message in the console:
Your pixel spacing is different along the x and y axes, make sure that you have not inversed the --image-width and the --image-height options
This might indicate an issue with swapped width and height dimensions, but I haven’t manually modified these parameters.
Command Used
I run the conversion using the following command:
./OrthancWSIDicomizer --openslide=./libopenslide.so “path/to/image/CMU-1.svs” --dataset=dataset.json --folder “path/to/output/image”
Questions
- Has anyone encountered this issue before?
- Is there a specific parameter to adjust to ensure the dimensions remain consistent between the original and converted images?
- Could the error message displayed before the conversion explain this difference in dimensions?
Thanks in advance for your help!
Lou-Anne