Automatic identification of RNCM position for TMA cores

[ngreenwald]

Is your feature request related to a problem? Please describe.
The current pipeline extracts whole TIFF images for each peak. For TMAs, however, the entire TMA is extracted at once. In order to FOV-level statistics, each core needs to be identified from the single extracted tif.

Describe the solution you'd like
A tool to automatically identify and crop out each TMA core in a given image. These coordinates can then be reused across all of the channels from a given TMA to produce FOV-specific crops.

One possibility would be looking into using Albert's TMA tiling tool to accomplish this. Another option would be building something ourselves to identify the RNCM postions.

Following this step, we'd need to actually crop out each FOV. This would likely be fairly straightforward with some thresholding/image processing, as the image that gets produced has tons of contrast:

[ngreenwald]

Based on the email chain, it sounds like we won't get additional information from Gray right away. In the meantime, Peggi suggested we look into the SDK:

Bruker has software development kits that we can use to extract the spatial data without the processing done by SCiLS. This is in the bruker library, search timstofsdk, raw access & api, and grab the ts-sdk which is the toolkit for the spatial library.

We should take a look. It's possible that this contains the exact coordinates of each ROI. If we can export this, we wouldn't need to do any image processing to generate the masks