Add function to extract ROIs metadata from ScanImage tiff

[alessandratrapani]

In the ScanImage .tif metadata there is a dictonary saved as text that the function extract_extra_metadata cannot parse, where information on the ROIs are stored. I implemented a new function call extract_roi_metadata (#272) that return this dict.

"RoiGroups":{
    "imagingRoiGroup":{# general imaging space metadata (only 2D)
        ...
        "centerXY": [-1.776356839e-16,0.1],
        "sizeXY": [13.5,13.5], # size of the FOV in micrometers
        "rotationDegrees": 0,
        "enable": 1,
        "pixelResolutionXY": [512,512],
        "pixelToRefTransform": [[0.0263671875,0,-6.763183594],
                                [0,0.263671875,-6.663183594],
                                [0,0,1]],
        "affine":  [[13.5,0,-6.75],
                    [0,13.5,-6.65],
                    [0,0,1]]
        ...
    },
    "integrationRoiGroup":{
        ...
        "rois": [
        {
          "ver": 1,
          "classname": "scanimage.mroi.Roi",
          "name": "ROI 1",
            ...
          "zs": 0,
          "scanfields": {
            ...
            "centerXY": [2.518066406,1.062402344],
            "sizeXY": [0.2373046875,0.2373046875],
            "rotationDegrees": 0,
            "enable": 1,
            "threshold": 100,
            "channel": 1,
            "processor": "cpu",
            "mask": [roi_pixel_mask 9x9 binary matrix],
            "affine": [[0.2373046875,0,2.518066406],
                       [0,0.2373046875,1.062402344],
                       [0,0,1]]
          },
          "discretePlaneMode": 1,
          "powers": null,
          "pzAdjust": null,
          "Lzs": null,
          "interlaceDecimation": null,
          "interlaceOffset": null,
          "enable": 1
        },
        ... # all the 368 ROIs
        ]
    } # imaging space metadata for each ROI
}

From imagingRoiGroup we can extract info on the dimension of the field of view (XY) from "sizeXY" and divide by the dimension to get the grid_spacing (z dimension missed)

In integrationRoiGroup we can extract information about the position of the rois: inv(pixelToRefTransform)*centerXY

[pauladkisson]

Thank you for catching this. It looks like there's a whole swath of metadata stored in a json-style string that we haven't been extracting.

However, I would prefer if this were integrated into the existing extract_extra_metadata function. That way the metadata function would return all the metadata in the 'system_metadata_string' AND the 'extra_metadata_string' (as you called them). Then you can add a section to grab the ROI groups in the parse_metadata function.

[pauladkisson]

LGTM! Thanks for implementing the changes.

[codecov]

Codecov Report

Attention: 8 lines in your changes are missing coverage. Please review.

Comparison is base (bd18e5b) 79.19% compared to head (44194b3) 79.00%.
Report is 1 commits behind head on main.

❗ Current head 44194b3 differs from pull request most recent head 1d385ee. Consider uploading reports for the commit 1d385ee to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #272      +/-   ##
==========================================
- Coverage   79.19%   79.00%   -0.20%     
==========================================
  Files          39       39              
  Lines        3057     3067      +10     
==========================================
+ Hits         2421     2423       +2     
- Misses        636      644       +8     

Flag	Coverage Δ
unittests	79.00% <20.00%> (-0.20%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files	Coverage Δ
...ctors/tiffimagingextractors/scanimagetiff_utils.py	87.87% <20.00%> (-12.13%)	⬇️