XNAT Machine Learning 1.8 - how to just use an already trained Modell -

[computerv...@gmail.com]

Hello XNAT Community,

maybe I am mistaken. I am not the Machine Learning expert.

1. But is it possible to just use a trained ready model in xnat for segmentation?

I installed all what I need and I am also ready to train models for myself, but I didnt find something in the dokumentation on how to just use a trained model. Lets say I have a new dataset from the clinic and want to segment/label tumors or something automaticly.

2. And second, I have the datasets to train my model. Now I trained my modell in xnat. How to I check with the testdatasets and find my dice scores?

Do I have a mistake of thinking?

If the answer is to complicated, maybe somebody has something to ready for beginners.

Thanks,

Ralf