when integrating lateralising only data (lat_not_loc), use subset of hemispheric GIFs already in all_gifs_combined

[thenineteen]

Only pd.concat(join=inner) (if reindexed) or Gif Parcellations that already exist in all_combined_gifs

this avoids excess addition of false localising values to irrelevant regions

related to the changes made in pull #167

[thenineteen]

there is no right way about this

this is the idea that we are integrating lateralisation and localisation data rather than querying them separately. And if we already have some localisation data even from other patients and publications, then it may be easier on the eye to see few brain regions light up and more accurate. OTherwise, the lateralisation which exceed localisation or the lateralisations without localisations result in entire hemispheres lighting up which can give the false impression of data suggesting e.g. occipital lobe resections resulting in seizure freedom from gelastic seizures.

[thenineteen]

This would have to be implemented for both lat_but_not_loc and lat_exceeds_loc both in Q_L

[thenineteen]

given the current SVT (3D slicer GUI) has a Visualisation tab with Min. 2D opacity setting set at 0.25 (makes the top 25% opaque) - I think I will leave this feature out for now, because:

• the lowest 75% of GIF segmentations are translucent in the current SVT, by default. This means lat_exceed_loc may generate novel GIF regions due to mapping hemispherically, but if these regions have few other datapoints, they will be translucent compared to other regions for which explicit loc data is available.
• leaving it open in case others want the feature in the future

[thenineteen]

this is implemented better in the global lateralisation option