sergivalverde / nicMSlesions

Easy multiple sclerosis white matter lesion segmentation using convolutional deep neural networks.
GNU General Public License v3.0
46 stars 21 forks source link

Inference segmentation from de baseline_2ch #19

Closed isabellameds closed 2 years ago

isabellameds commented 3 years ago

Dear Sergi,

I have a question regarding the inference process, is it possible to predict the lesions using just your pretrained model? Or how to predict using only a full trained model? I tried to do that today, but the results are weird, I think I may be writing something wrong in the configuration file. I used the following parameters regarding the model:

[train] full_train = True pretrained_model = None

[model] name = baseline_2ch pretrained = None train_split = 0.25 max_epochs = 400 patience = 50 batch_size = 128 net_verbose = 1 gpu_number = 0

Is there something wrong?

Kind regards, Isabella