search

xmengli / H-DenseUNet

TMI 2018. H-DenseUNet: Hybrid Densely Connected UNet for Liver and Tumor Segmentation from CT Volumes
526 stars 161 forks source link

About generate livermask #114

Closed Ostarsion closed 2 years ago

Ostarsion commented 2 years ago

I notice that you provide us 70 generated test CT liver mask ,now i wanna try this method on my own data, but I find this code don't have a part to generate liver mask. there is a function called 'predict_window_mulgpu' in func.py seems to do that thing ,but it also need a model to support. Is it also 'model_best.hdf5' ? if not ,how could i get that model ? If possible， please reply ，Thank you so much.

Ostarsion commented 2 years ago

I find this step is not so important for training ,its just for speeding up the test part ,anyway ty.