Using mask for patching

[delianaramos]

I'm using Pathopatcher since it's integrated in CellVIT++ and want to use my own tissue mask for the patching and further inference.

I changed the code in order to use a json, but in inference the result is the same as if i would have not use it, despite the mask it's correctly loaded.

Do you know in which part I'm wrong?

Another issue I have it is when normalizing patches I got
[WARNING] - numpy.linalg.LinAlgError: Eigenvalues did not converge. The normalization could not be performed, try using a higher level with more patches.
But don´t know which level it does refer to...

Thanks
Ana