def build_model():
l_in = nn.layers.InputLayer((None, nslices, 30) + patch_size)
l_in_slice_mask = nn.layers.InputLayer((None, nslices))
l_in_slice_location = nn.layers.InputLayer((None, nslices, 1))
l_in_sex_age = nn.layers.InputLayer((None, 2))
l_ins = [l_in, l_in_slice_mask, l_in_slice_location, l_in_sex_age]
l_in_rshp = nn.layers.ReshapeLayer(l_in, (-1, 30) + patch_size) # (batch_size*nslices, 30, h,w)
submodel = subconfig().build_model(l_in_rshp)
# ------------------ systole
l_mu0 = submodel.mu_layers[0]
l_sigma0 = submodel.sigma_layers[0]
l_mu0 = nn.layers.ReshapeLayer(l_mu0, (-1, nslices, [1]))
l_sigma0 = nn.layers.ReshapeLayer(l_sigma0, (-1, nslices, [1]))
l_volume_mu_sigma0 = nn_heart.JeroenLayer([nn.layers.flatten(l_mu0), nn.layers.flatten(l_sigma0),
l_in_slice_mask, nn.layers.flatten(l_in_slice_location)],
trainable_scale=False)
l_volume_mu0 = nn.layers.SliceLayer(l_volume_mu_sigma0, indices=0, axis=-1)
l_volume_sigma0 = nn.layers.SliceLayer(l_volume_mu_sigma0, indices=1, axis=-1)
l_cdf0 = nn_heart.NormalCDFLayer(nn.layers.reshape(l_volume_mu0, ([0], 1)),
nn.layers.reshape(l_volume_sigma0, ([0], 1)))
# ------------------ diastole
l_mu1 = submodel.mu_layers[1]
l_sigma1 = submodel.sigma_layers[1]
l_mu1 = nn.layers.ReshapeLayer(l_mu1, (-1, nslices, [1]))
l_sigma1 = nn.layers.ReshapeLayer(l_sigma1, (-1, nslices, [1]))
l_volume_mu_sigma1 = nn_heart.JeroenLayer([nn.layers.flatten(l_mu1), nn.layers.flatten(l_sigma1),
l_in_slice_mask, nn.layers.flatten(l_in_slice_location)],
trainable_scale=False)
l_volume_mu1 = nn.layers.SliceLayer(l_volume_mu_sigma1, indices=0, axis=-1)
l_volume_sigma1 = nn.layers.SliceLayer(l_volume_mu_sigma1, indices=1, axis=-1)
l_cdf1 = nn_heart.NormalCDFLayer(nn.layers.reshape(l_volume_mu1, ([0], 1)),
nn.layers.reshape(l_volume_sigma1, ([0], 1)))
l_outs = [l_cdf0, l_cdf1]
l_top = nn.layers.MergeLayer(l_outs)
l_target_mu0 = nn.layers.InputLayer((None, 1))
l_target_mu1 = nn.layers.InputLayer((None, 1))
l_targets = [l_target_mu0, l_target_mu1]
train_params = nn.layers.get_all_params(l_top)
test_layes = [l_volume_mu_sigma0, l_volume_mu_sigma1]
return namedtuple('Model', ['l_ins', 'l_outs', 'l_targets', 'l_top', 'train_params', 'submodel', 'test_layers'])(
l_ins, l_outs,
l_targets,
l_top,
train_params,
submodel, test_layes)
def build_model():
l_in = nn.layers.InputLayer((None, nslices, 30) + patch_size)
l_in_slice_mask = nn.layers.InputLayer((None, nslices))
l_in_slice_location = nn.layers.InputLayer((None, nslices, 1))
l_in_sex_age = nn.layers.InputLayer((None, 2))
l_ins = [l_in, l_in_slice_mask, l_in_slice_location, l_in_sex_age]
l_in_rshp = nn.layers.ReshapeLayer(l_in, (-1, 30) + patch_size) # (batch_size*nslices, 30, h,w)
submodel = subconfig().build_model(l_in_rshp)
# ------------------ systole
l_mu0 = submodel.mu_layers[0]
l_sigma0 = submodel.sigma_layers[0]
l_mu0 = nn.layers.ReshapeLayer(l_mu0, (-1, nslices, [1]))
l_sigma0 = nn.layers.ReshapeLayer(l_sigma0, (-1, nslices, [1]))
l_volume_mu_sigma0 = nn_heart.JeroenLayer([nn.layers.flatten(l_mu0), nn.layers.flatten(l_sigma0),
l_in_slice_mask, nn.layers.flatten(l_in_slice_location)],
trainable_scale=False)
l_volume_mu0 = nn.layers.SliceLayer(l_volume_mu_sigma0, indices=0, axis=-1)
l_volume_sigma0 = nn.layers.SliceLayer(l_volume_mu_sigma0, indices=1, axis=-1)
l_cdf0 = nn_heart.NormalCDFLayer(nn.layers.reshape(l_volume_mu0, ([0], 1)),
nn.layers.reshape(l_volume_sigma0, ([0], 1)))
# ------------------ diastole
l_mu1 = submodel.mu_layers[1]
l_sigma1 = submodel.sigma_layers[1]
l_mu1 = nn.layers.ReshapeLayer(l_mu1, (-1, nslices, [1]))
l_sigma1 = nn.layers.ReshapeLayer(l_sigma1, (-1, nslices, [1]))
l_volume_mu_sigma1 = nn_heart.JeroenLayer([nn.layers.flatten(l_mu1), nn.layers.flatten(l_sigma1),
l_in_slice_mask, nn.layers.flatten(l_in_slice_location)],
trainable_scale=False)
l_volume_mu1 = nn.layers.SliceLayer(l_volume_mu_sigma1, indices=0, axis=-1)
l_volume_sigma1 = nn.layers.SliceLayer(l_volume_mu_sigma1, indices=1, axis=-1)
l_cdf1 = nn_heart.NormalCDFLayer(nn.layers.reshape(l_volume_mu1, ([0], 1)),
nn.layers.reshape(l_volume_sigma1, ([0], 1)))
l_outs = [l_cdf0, l_cdf1]
l_top = nn.layers.MergeLayer(l_outs)
l_target_mu0 = nn.layers.InputLayer((None, 1))
l_target_mu1 = nn.layers.InputLayer((None, 1))
l_targets = [l_target_mu0, l_target_mu1]
train_params = nn.layers.get_all_params(l_top)
test_layes = [l_volume_mu_sigma0, l_volume_mu_sigma1]
return namedtuple('Model', ['l_ins', 'l_outs', 'l_targets', 'l_top', 'train_params', 'submodel', 'test_layers'])(
l_ins, l_outs,
l_targets,
l_top,
train_params,
submodel, test_layes)
from collections import namedtuple
import lasagne as nn
import data_iterators
import numpy as np
import theano.tensor as T
import nn_heart
from configuration import subconfig
import utils_heart
from pathfinder import PKL_TRAIN_DATA_PATH, TRAIN_LABELS_PATH, PKL_VALIDATE_DATA_PATH
import utils
caching = None
restart_from_save = None
rng = subconfig().rng
patch_size = subconfig().patch_size
train_transformation_params = subconfig().train_transformation_params
valid_transformation_params = subconfig().valid_transformation_params
test_transformation_params = subconfig().test_transformation_params
batch_size = 8
nbatches_chunk = 2
chunk_size = batch_size * nbatches_chunk
train_valid_ids = utils.get_train_valid_split(PKL_TRAIN_DATA_PATH)
train_data_iterator = data_iterators.PatientsDataGenerator(data_path=PKL_TRAIN_DATA_PATH,
batch_size=chunk_size,
transform_params=train_transformation_params,
patient_ids=train_valid_ids['train'],
labels_path=TRAIN_LABELS_PATH,
from collections import namedtuple
import lasagne as nn
import data_iterators
import numpy as np
import theano.tensor as T
import nn_heart
from configuration import subconfig
import utils_heart
import utils
from pathfinder import PKL_TRAIN_DATA_PATH, TRAIN_LABELS_PATH, PKL_VALIDATE_DATA_PATH
caching = None
restart_from_save = None
rng = subconfig().rng
patch_size = subconfig().patch_size
train_transformation_params = subconfig().train_transformation_params
valid_transformation_params = subconfig().valid_transformation_params
test_transformation_params = subconfig().test_transformation_params
data_prep_fun = subconfig().data_prep_fun
batch_size = 8
nbatches_chunk = 2
chunk_size = batch_size * nbatches_chunk
train_valid_ids = utils.get_train_valid_split(PKL_TRAIN_DATA_PATH)
train_data_iterator = data_iterators.PatientsDataGenerator(data_path=PKL_TRAIN_DATA_PATH,
batch_size=chunk_size,
transform_params=train_transformation_params,
patient_ids=train_valid_ids['train'],
