OpenHI coordinate system

OpenHI uses specific terms to address images during the processing.

../_images/coordinate_system1.png

OpenHI coordinate system

Refer to a region (area) in WSI

Regions in WSI are referred to as viewing area in OpenHI. This is the green box in the image above. There are two parameters to describe this box.

1. Top-left coordinate

The top-left coordinate is an anchor specifying where are you referring to in WSI (which is usually very big). In OpenHI, top-left coordinates came in pair (x, y). The minimum value is (0, 0) and the maximum value depends on the size of particular WSI (get WSI size) which equals to the bottom right or the width (dashed blue line) and height (dashed purple line) of the WSI.

2. Region size

Depends on how large the area you want to view. Region size or (sometimes called viewing_size) will specify the area you wish to view, thus this is called viewing area. The parameter also came in pair (width, height) which came in int.

Returning image

Users can specify a very large area in the WSI. Thus we need to specify the viewer size (blue rectangle) as well since we cannot just return very large image array (that would take up a large chunk of machine’s memory and could crash the program in case someone request to view 100,000 by 100,000 area!). In the process of reading WSI files, down sampling is part of the job. In short, users must define how large they want the returning image is going to be so that the library can fetch the image at the correct level (it’s a pyramidal image).

The viewer_size consist of (width, height) pair. It would be better if the size ratio match the viewing_size (see region size) so the output image would be squished. For example, when the down sampling factor (DSF) is 2:

viewing_size = (1000, 2000)
viewer_size = (500, 1000)

Example usage:

openhi.slide.Slide.read_region()