Transforms such as macenko normalizer are now done on each slide separately.
It might be preferred to fit it on the whole dataset, and then do the normalization with that.
However, im not sure yet how to accomplish this in the current setup of the framework.
Transforms such as macenko normalizer are now done on each slide separately.
It might be preferred to fit it on the whole dataset, and then do the normalization with that.
However, im not sure yet how to accomplish this in the current setup of the framework.