def __init__(
self,
output_dir: str = "./",
output_postfix: str = "trans",
output_ext: str = ".nii.gz",
resample: bool = True,
mode: Union[GridSampleMode, InterpolateMode, str] = "nearest",
padding_mode: Union[GridSamplePadMode, str] = GridSamplePadMode.BORDER,
scale: Optional[int] = None,
dtype: DtypeLike = np.float64,
output_dtype: DtypeLike = np.float32,
save_batch: bool = False,
squeeze_end_dims: bool = True,
data_root_dir: str = "",
print_log: bool = True,
) -> None:
self.saver: Union[NiftiSaver, PNGSaver]
if output_ext in (".nii.gz", ".nii"):
self.saver = NiftiSaver(
output_dir=output_dir,
output_postfix=output_postfix,
output_ext=output_ext,
resample=resample,
mode=GridSampleMode(mode),
padding_mode=padding_mode,
dtype=dtype,
output_dtype=output_dtype,
squeeze_end_dims=squeeze_end_dims,
data_root_dir=data_root_dir,
print_log=print_log,
)
elif output_ext == ".png":
self.saver = PNGSaver(
output_dir=output_dir,
output_postfix=output_postfix,
output_ext=output_ext,
resample=resample,
mode=InterpolateMode(mode),
scale=scale,
data_root_dir=data_root_dir,
print_log=print_log,
)
else:
raise ValueError(f"unsupported output extension: {output_ext}.")
self.save_batch = save_batch
#plt.show()
#fig2.savefig('Pancreas_Plot.png')
"""## Check best model output with the input image and label"""
"""## Makes the Inferences """
out_dir = "//home//imoreira//Data//Kidneys_Best_Model"
#out_dir = "C:\\Users\\isasi\\Downloads\\Pancreas_Best_Model"
model.load_state_dict(
torch.load(os.path.join(out_dir, "best_metric_model.pth")))
model.eval()
with torch.no_grad():
#saver = NiftiSaver(output_dir='C:\\Users\\isasi\\Downloads\\Pancreas_Segs_Out')
saver = NiftiSaver(output_dir='//home//imoreira//Kidneys_Segs_Out',
output_postfix="seg_kidneys",
output_ext=".nii.gz",
mode="nearest",
padding_mode="zeros")
for i, train_inf_data in enumerate(train_inf_loader):
#for test_data in test_loader:
train_inf_images = train_inf_data["image"].to(device)
roi_size = (96, 96, 96)
sw_batch_size = 4
val_outputs = sliding_window_inference(train_inf_images,
roi_size,
sw_batch_size,
model,
overlap=0.8)
# val_outputs = torch.squeeze(val_outputs, dim=1)
plt.show()
fig2.savefig('Lungs_Plot.png')
"""## Check best model output with the input image and label"""
"""## Makes the Inferences """
out_dir = "//home//imoreira//Data//Output"
#out_dir = "C:\\Users\\isasi\\Downloads\\Output"
model.load_state_dict(
torch.load(os.path.join(out_dir, "best_metric_model.pth")))
model.eval()
with torch.no_grad():
#saver = NiftiSaver(output_dir='C:\\Users\\isasi\\Downloads\\Segmentations')
saver = NiftiSaver(output_dir='//home//imoreira//Segmentations',
output_postfix="seg_lungs",
output_ext=".nii.gz",
mode="nearest",
padding_mode="zeros")
for i, test_data in enumerate(test_loader):
test_images = test_data["image"].to(device)
roi_size = (160, 160, 160)
sw_batch_size = 4
val_outputs_1 = sliding_window_inference(test_images, roi_size,
sw_batch_size, model)
val_outputs_2 = sliding_window_inference(test_images, roi_size,
sw_batch_size, model)
val_outputs_1 = val_outputs_1.argmax(dim=1, keepdim=True)
val_outputs_2 = val_outputs_2.argmax(dim=1, keepdim=True)
# plt.show()
# fig2.savefig('Training_Plot.png')
"""## Check best model output with the input image and label"""
"""## Makes the Inferences """
###
out_dir = "//home//imoreira//Data//MO_Best_Model"
#out_dir = "C:\\Users\\isasi\\Downloads\\Bladder_Best_Model"
model.load_state_dict(
torch.load(os.path.join(out_dir, "best_metric_model.pth")))
model.eval()
with torch.no_grad():
#saver = NiftiSaver(output_dir='C:\\Users\\isasi\\Downloads\\Bladder_Segs_Out')
saver = NiftiSaver(
output_dir='//home//imoreira//Segs_Out//',
#output_dir='C:\\Users\\isasi\\Downloads\\Segs_Out',
output_postfix="seg",
output_ext=".nii.gz",
mode="nearest",
padding_mode="zeros")
for i, test_data in enumerate(test_loader):
#for test_data in test_loader:
test_images = test_data["image"].to(device)
roi_size = (96, 96, 96)
sw_batch_size = 4
val_outputs = sliding_window_inference(test_images,
roi_size,
sw_batch_size,
model,
overlap=0.8)
val_outputs = val_outputs.argmax(dim=1, keepdim=True)
strides=(2, 2, 2, 2),
num_res_units=2,
norm=Norm.BATCH,
).to(device)
loss_function = DiceLoss(to_onehot_y=True, softmax=False, sigmoid=True)
optimizer = torch.optim.Adam(model.parameters(), 1e-4)
"""## Makes the Inferences """
model.load_state_dict(
torch.load(os.path.join(out_dir, "best_metric_model.pth")))
model.eval()
with torch.no_grad():
#saver = NiftiSaver(output_dir='C:\\Users\\isasi\\Downloads\\Bladder_Segs_Out')
saver = NiftiSaver(output_dir='//home//imoreira//Segs_Adam',
output_postfix="seg",
output_ext=".nii.gz",
mode="nearest",
padding_mode="zeros")
for test_data in test_loader:
test_images = test_data["image"].to(device)
roi_size = (160, 160, 160)
sw_batch_size = 4
val_outputs = sliding_window_inference(
test_images,
roi_size,
sw_batch_size,
model #overlap=0.8
)
val_outputs = val_outputs.argmax(dim=1, keepdim=True)
val_outputs = val_outputs.cpu().clone().numpy()
val_outputs = val_outputs.astype(np.int)
