def get_arguments():
parser = argparse.ArgumentParser()
parser.add_argument('--batchSz', type=int, default=4)
parser.add_argument('--dataset_name', type=str, default="iseg2017")
parser.add_argument('--dim', nargs="+", type=int, default=(32,32,32))
parser.add_argument('--nEpochs', type=int, default=250)
parser.add_argument('--classes', type=int, default=4)
parser.add_argument('--samples_train', type=int, default=1)
parser.add_argument('--samples_val', type=int, default=1)
parser.add_argument('--split', type=float, default=0.8)
parser.add_argument('--inChannels', type=int, default=2)
parser.add_argument('--inModalities', type=int, default=2)
parser.add_argument('--fold_id', default='1', type=str, help='Select subject for fold validation')
parser.add_argument('--lr', default=1e-2, type=float,
help='learning rate (default: 1e-3)')
parser.add_argument('--cuda', action='store_true', default=True)
parser.add_argument('--resume', default='', type=str, metavar='PATH',
help='path to latest checkpoint (default: none)')
parser.add_argument('--model', type=str, default='UNET3D',
choices=('VNET', 'VNET2', 'UNET3D', 'DENSENET1', 'DENSENET2', 'DENSENET3', 'HYPERDENSENET'))
parser.add_argument('--opt', type=str, default='sgd',
choices=('sgd', 'adam', 'rmsprop'))
parser.add_argument('--pretrained',
default='../saved_models/UNET3D_checkpoints/UNET3D_24_04___14_27_iseg2017_/UNET3D_24_04___14_27_iseg2017__last_epoch.pth',
type=str, metavar='PATH',
help='path to pretrained model')
args = parser.parse_args()
args.save = '../inference_checkpoints/' + args.model + '_checkpoints/' + args.model + '_{}_{}_'.format(
utils.datestr(), args.dataset_name)
args.tb_log_dir = '../runs/'
return args
def get_arguments():
parser = argparse.ArgumentParser()
parser.add_argument('--batchSz', type=int, default=2)
parser.add_argument('--dataset_name', type=str, default="brats2019")
parser.add_argument('--dim', nargs="+", type=int, default=(64, 64, 64))
parser.add_argument('--nEpochs', type=int, default=100)
parser.add_argument('--classes', type=int, default=5)
parser.add_argument('--samples_train', type=int, default=100)
parser.add_argument('--samples_val', type=int, default=100)
parser.add_argument('--inChannels', type=int, default=4)
parser.add_argument('--inModalities', type=int, default=4)
parser.add_argument('--split', default=0.8, type=float, help='Select percentage of training data(default: 0.8)')
parser.add_argument('--lr', default=1e-3, type=float,
help='learning rate (default: 1e-3)')
parser.add_argument('--cuda', action='store_true', default=True)
parser.add_argument('--resume', default='', type=str, metavar='PATH',
help='path to latest checkpoint (default: none)')
parser.add_argument('--model', type=str, default='UNET3D',
choices=('VNET', 'VNET2', 'UNET3D', 'DENSENET1', 'DENSENET2', 'DENSENET3', 'HYPERDENSENET'))
parser.add_argument('--opt', type=str, default='sgd',
choices=('sgd', 'adam', 'rmsprop'))
parser.add_argument('--log_dir', type=str,
default='../runs/')
args = parser.parse_args()
args.save = '../saved_models/' + args.model + '_checkpoints/' + args.model + '_{}_{}_'.format(
utils.datestr(), args.dataset_name)
return args
def main():
args = get_arguments()
utils.make_dirs(args.save)
name_model = args.model + "_" + args.dataset_name + "_" + utils.datestr()
# TODO visual3D_temp.Basewriter package
writer = SummaryWriter(log_dir='../runs/' + name_model, comment=name_model)
training_generator, val_generator, full_volume, affine = medical_loaders.generate_datasets(
args, path='.././datasets')
model, optimizer = medzoo.create_model(args)
criterion = DiceLoss(classes=11, skip_index_after=args.classes)
if args.cuda:
torch.cuda.manual_seed(seed)
model = model.cuda()
print("Model transferred in GPU.....")
print("START TRAINING...")
for epoch in range(1, args.nEpochs + 1):
train_stats = train.train_dice(args, epoch, model, training_generator,
optimizer, criterion)
val_stats = train.test_dice(args, epoch, model, val_generator,
criterion)
#old
utils.write_train_val_score(writer, epoch, train_stats, val_stats)
model.save_checkpoint(args.save,
epoch,
val_stats[0],
optimizer=optimizer)
def get_arguments():
parser = argparse.ArgumentParser()
parser.add_argument('--batchSz', type=int, default=4)
parser.add_argument('--dataset_name', type=str, default="miccai2019")
parser.add_argument('--nEpochs', type=int, default=100)
parser.add_argument('--dim', nargs="+", type=int, default=(256, 256))
parser.add_argument('--classes', type=int, default=7)
parser.add_argument('--inChannels', type=int, default=3)
parser.add_argument('--inModalities', type=int, default=1)
parser.add_argument('--samples_train', type=int, default=100)
parser.add_argument('--samples_val', type=int, default=10)
parser.add_argument('--fold_id', default='1', type=str, help='Select subject for fold validation')
parser.add_argument('--lr', default=1e-3, type=float,
help='learning rate (default: 1e-3)')
parser.add_argument('--cuda', action='store_true', default=True)
parser.add_argument('--resume', default='', type=str, metavar='PATH',
help='path to latest checkpoint (default: none)')
parser.add_argument('--model', type=str, default='UNET2D',
choices=(
'VNET', 'VNET2', 'UNET3D', 'DENSENET1', 'DENSENET2', 'DENSENET3', 'HYPERDENSENET',
"UNET2D"))
parser.add_argument('--opt', type=str, default='sgd',
choices=('sgd', 'adam', 'rmsprop'))
args = parser.parse_args()
args.save = '../saved_models/' + args.model + '_checkpoints/' + args.model + '_{}_{}_'.format(
utils.datestr(), args.dataset_name)
return args
def main():
args = get_arguments()
seed = 1777777
torch.manual_seed(seed)
if args.cuda:
torch.cuda.manual_seed(seed)
np.random.seed(seed)
cudnn.deterministic = True
cudnn.benchmark = True
utils.make_dirs(args.save)
name_model = args.model + "_" + args.dataset_name + "_" + utils.datestr()
# TODO visual3D_temp.Basewriter package
writer = SummaryWriter(log_dir='../runs/' + name_model, comment=name_model)
training_generator, val_generator, full_volume, affine = medical_loaders.generate_datasets(
args, path='.././datasets')
model, optimizer = medzoo.create_model(args)
if args.cuda:
model = model.cuda()
print("Model transferred in GPU.....")
print("START TRAINING...")
for epoch in range(1, args.nEpochs + 1):
train(args, model, training_generator, optimizer, epoch, writer)
val_metrics, confusion_matrix = validation(args, model, val_generator,
epoch, writer)
def main():
args = get_arguments()
os.environ["CUDA_VISIBLE_DEVICES"] = "0,2"
## FOR REPRODUCIBILITY OF RESULTS
seed = 1777777
utils.reproducibility(args, seed)
utils.make_dirs(args.save)
name_model = args.model + "_" + args.dataset_name + "_" + utils.datestr()
# TODO visual3D_temp.Basewriter package
writer = SummaryWriter(log_dir='../runs/' + name_model, comment=name_model)
training_generator, val_generator, full_volume, affine = medical_loaders.generate_datasets(
args, path='.././datasets')
model, optimizer = medzoo.create_model(args)
if args.cuda:
model = model.cuda()
print("Model transferred in GPU.....")
print("START TRAINING...")
for epoch in range(1, args.nEpochs + 1):
train(args, model, training_generator, optimizer, epoch, writer)
val_metrics, confusion_matrix = validation(args, model, val_generator,
epoch, writer)
def get_arguments():
parser = argparse.ArgumentParser()
parser.add_argument('--batchSz', type=int, default=4)
parser.add_argument('--dataset_name', type=str, default="mrbrains4")
parser.add_argument('--dataset_path', type=str, default="datasets")
parser.add_argument('--dim', nargs="+", type=int, default=(128, 128, 48))
parser.add_argument('--nEpochs', type=int, default=200)
parser.add_argument('--inChannels', type=int, default=3)
parser.add_argument('--inModalities', type=int, default=3)
parser.add_argument('--terminal_show_freq', default=50)
parser.add_argument('--samples_train', type=int, default=10)
parser.add_argument('--samples_val', type=int, default=10)
parser.add_argument('--classes', type=int, default=2)
parser.add_argument('--threshold', default=0.1, type=float)
parser.add_argument('--augmentation',
default='no',
type=str,
help='Tensor normalization: options max, mean, global')
parser.add_argument('--normalization',
default='global_mean',
type=str,
help='Tensor normalization: options max, mean, global')
parser.add_argument('--resume',
default='',
type=str,
metavar='PATH',
help='path to latest checkpoint (default: none)')
parser.add_argument(
'--split',
default=0.9,
type=float,
help='Select percentage of training data(default: 0.8)')
parser.add_argument('--lr',
default=1e-3,
type=float,
help='learning rate (default: 1e-3)')
parser.add_argument('--cuda', action='store_true', default=False)
parser.add_argument('--model',
type=str,
default='UNET3D',
choices=('VNET', 'VNET2', 'UNET3D', 'DENSENET1',
'DENSENET2', 'DENSENET3', 'HYPERDENSENET'))
parser.add_argument('--opt',
type=str,
default='sgd',
choices=('sgd', 'adam', 'rmsprop'))
parser.add_argument('--log_dir', type=str, default='../runs/')
args = parser.parse_args()
args.save = '../saved_models/' + args.model + '_checkpoints/' + args.model + '_{}_{}_'.format(
utils.datestr(), args.dataset_name)
return args
def get_arguments():
parser = argparse.ArgumentParser()
parser.add_argument('--batchSz', type=int, default=2)
parser.add_argument('--dataset_name', type=str, default="iseg2019")
parser.add_argument('--dim', nargs="+", type=int, default=(32, 32, 32))
parser.add_argument('--nEpochs', type=int, default=200)
parser.add_argument('--classes', type=int, default=4)
parser.add_argument('--samples_train', type=int, default=10)
parser.add_argument('--samples_val', type=int, default=10)
parser.add_argument('--inChannels', type=int, default=2)
parser.add_argument('--inModalities', type=int, default=2)
parser.add_argument('--threshold', default=0.0001, type=float)
parser.add_argument('--terminal_show_freq', default=50)
parser.add_argument('--augmentation', action='store_true', default=True)
parser.add_argument('--normalization',
default='full_volume_mean',
type=str,
help='Tensor normalization: options ,max_min,',
choices=('max_min', 'full_volume_mean', 'brats', 'max',
'mean'))
parser.add_argument('--fold_id',
default='1',
type=str,
help='Select subject for fold validation')
parser.add_argument('--lr',
default=1e-3,
type=float,
help='learning rate (default: 1e-3)')
parser.add_argument(
'--split',
default=0.8,
type=float,
help='Select percentage of training data(default: 0.8)')
parser.add_argument('--cuda', action='store_true', default=True)
parser.add_argument('--loadData', default=True)
parser.add_argument('--resume',
default='',
type=str,
metavar='PATH',
help='path to latest checkpoint (default: none)')
parser.add_argument('--model',
type=str,
default='UNET3D',
choices=('VNET', 'VNET2', 'UNET3D', 'DENSENET1',
'DENSENET2', 'DENSENET3', 'HYPERDENSENET'))
parser.add_argument('--opt',
type=str,
default='sgd',
choices=('sgd', 'adam', 'rmsprop'))
parser.add_argument('--log_dir', type=str, default='../runs/')
args = parser.parse_args()
args.save = '../saved_models/' + args.model + '_checkpoints/' + args.model + '_{}_{}_'.format(
utils.datestr(), args.dataset_name)
return args
def get_arguments():
parser = argparse.ArgumentParser()
parser.add_argument('--batchSz', type=int, default=1)
parser.add_argument('--dataset_name', type=str, default="ribfrac")
parser.add_argument('--dim', nargs="+", type=int, default=(64, 64, 64))
parser.add_argument('--nEpochs', type=int, default=250)
parser.add_argument('--classes', type=int, default=6)
parser.add_argument('--samples_train', type=int, default=1)
parser.add_argument('--samples_val', type=int, default=1)
parser.add_argument('--split', type=float, default=0.8)
parser.add_argument('--inChannels', type=int, default=1)
parser.add_argument('--inModalities', type=int, default=1)
parser.add_argument('--fold_id', default='1', type=str, help='Select subject for fold validation')
parser.add_argument('--lr', default=1e-2, type=float,
help='learning rate (default: 1e-3)')
parser.add_argument('--cuda', action='store_true', default=True)
parser.add_argument('--resume', default='', type=str, metavar='PATH',
help='path to latest checkpoint (default: none)')
parser.add_argument('--model', type=str, default='UNET3D',
choices=('VNET', 'VNET2', 'UNET3D', 'DENSENET1', 'DENSENET2', 'DENSENET3', 'HYPERDENSENET','DENSEVOXELNET'))
parser.add_argument('--opt', type=str, default='sgd',
choices=('sgd', 'adam', 'rmsprop'))
parser.add_argument('--loadData', default=False)
# params need to be set for test
parser.add_argument('--pretrained',
# default='/data/hejy/MedicalZooPytorch/saved_models/UNET3D_checkpoints/UNET3D_64x64x48_0.1_weight0.1/UNET3D_64x64x48_0.1_weight0.1_BEST.pth',
# default = '/data/hejy/MedicalZooPytorch/saved_models/UNET3D_checkpoints/UNET3D_128x128x48_thresh0.1_sample1200_epoch600_fgbg/UNET3D_128x128x48_thresh0.1_sample1200_epoch600_fgbg_BEST.pth',
# default='/data/hejy/MedicalZooPytorch/saved_models/UNET3D_checkpoints/UNET3D_128x128x48_thresh0.3_weight0.01_sample400/UNET3D_128x128x48_thresh0.3_weight0.01_sample400_BEST.pth',
# default='/data/hejy/MedicalZooPytorch/saved_models/UNET3D_checkpoints/UNET3D_128x128x48_thresh0.3_weight0.01_sample400_epoch600_test_val+/UNET3D_128x128x48_thresh0.3_weight0.01_sample400_epoch600_test_val+_BEST.pth',
# default='/data/hejy/MedicalZooPytorch/saved_models/DENSEVOXELNET_checkpoints/DENSEVOXELNET_128x128x48_thresh0.1_weight1_sample400_epoch600_augdata/DENSEVOXELNET_128x128x48_thresh0.1_weight1_sample400_epoch600_augdata_BEST.pth',
# default='/data/hejy/MedicalZooPytorch/saved_models/UNET3D_checkpoints/UNET3D_128x128x48_thresh0.1_weight1_sample400_epoch600_augdata/UNET3D_128x128x48_thresh0.1_weight1_sample400_epoch600_augdata_BEST.pth',
# default='/data/hejy/MedicalZooPytorch/saved_models/UNET3D_checkpoints/UNET3D_128x128x48_thresh0.1_sample1200_epoch600_fgbg/UNET3D_128x128x48_thresh0.1_sample1200_epoch600_fgbg_last_epoch.pth',
# default = '/data/hejy/MedicalZooPytorch/saved_models/UNET3D_checkpoints/UNET3D_196x196x96_thresh0.1_weight1_sample400/UNET3D_196x196x96_thresh0.1_weight1_sample400_BEST.pth',
default = '/data/hejy/MedicalZooPytorch_copy/saved_models/UNET3D_checkpoints/UNET3D_256x256x256_thresh0.8_weight1_sample1200/UNET3D_256x256x256_thresh0.8_weight1_sample1200_BEST.pth',
# default = '/data/hejy/MedicalZooPytorch_2cls/saved_models/UNET3D_checkpoints/2cls_UNET3D_512x512x96_thresh0.6_weight1_sample200/2cls_UNET3D_512x512x96_thresh0.6_weight1_sample200_BEST.pth',
# default='/data/hejy/MedicalZooPytorch/saved_models/UNET3D_checkpoints/UNET3D_128x128x48_thresh0.3_sample400_epoch600_test_val+_wce/UNET3D_128x128x48_thresh0.3_sample400_epoch600_test_val+_wce_BEST.pth',
type=str, metavar='PATH',
help='path to pretrained model')
parser.add_argument('--nii_save_path', default='/data/hejy/MedicalZooPytorch_copy/saved_nii')
parser.add_argument('--test_path', default="/data/beijing/dataset/MICCAI-RibFrac2020/ribfrac-val-images")
# parser.add_argument('--test_path', default="/data/hejy/MedicalZooPytorch/datasets/MICCAI_2020_ribfrac/org/ribfrac-val-images")
args = parser.parse_args()
args.save = '/data/hejy/MedicalZooPytorch_copy/inference_checkpoints/' + args.model + '_checkpoints/' + args.model + '_{}_{}_'.format(
utils.datestr(), args.dataset_name)
args.tb_log_dir = '../runs/'
return args
def __init__(self, args):
name_model = args.log_dir + args.model + "_" + args.dataset_name + "_" + utils.datestr(
)
self.writer = SummaryWriter(log_dir=args.log_dir + name_model,
comment=name_model)
utils.make_dirs(args.save)
self.csv_train, self.csv_val = self.create_stats_files(args.save)
self.dataset_name = args.dataset_name
self.classes = args.classes
self.label_names = dict_class_names[args.dataset_name]
self.data = self.create_data_structure()
def get_arguments():
parser = argparse.ArgumentParser()
parser.add_argument('--batchSz', type=int, default=1)
parser.add_argument('--dataset_name', type=str, default="iseg2017")
parser.add_argument('--dim', nargs="+", type=int, default=(32, 32, 32))
parser.add_argument('--nEpochs', type=int, default=250)
parser.add_argument('--classes', type=int, default=4)
parser.add_argument('--samples_train', type=int, default=10)
parser.add_argument('--samples_val', type=int, default=10)
parser.add_argument('--inChannels', type=int, default=2)
parser.add_argument('--inModalities', type=int, default=2)
parser.add_argument('--fold_id',
default='1',
type=str,
help='Select subject for fold validation')
parser.add_argument('--lr',
default=1e-3,
type=float,
help='learning rate (default: 1e-3)')
parser.add_argument('--cuda', action='store_true', default=False)
parser.add_argument('--resume',
default='',
type=str,
metavar='PATH',
help='path to latest checkpoint (default: none)')
parser.add_argument('--model',
type=str,
default='DENSEVOXELNET',
choices=('VNET', 'VNET2', 'UNET3D', 'DENSENET1',
'DENSENET2', 'DENSENET3', 'HYPERDENSENET'))
parser.add_argument('--opt',
type=str,
default='sgd',
choices=('sgd', 'adam', 'rmsprop'))
parser.add_argument(
'--log_dir',
type=str,
default=
'/home/papastrat/PycharmProjects/MedicalZooPytorch/runs/lias_test_run')
args = parser.parse_args()
args.save = '../saved_models/' + args.model + '_checkpoints/' + args.model + '_{}_{}_'.format(
datestr(), args.dataset_name)
return args
def main():
args = get_arguments()
utils.make_dirs(args.save)
train_f, val_f = utils.create_stats_files(args.save)
name_model = args.model + "_" + args.dataset_name + "_" + utils.datestr()
writer = SummaryWriter(log_dir='../runs/' + name_model, comment=name_model)
best_prec1 = 100.
training_generator, val_generator, full_volume, affine = medical_loaders.generate_datasets(
args, path='.././datasets')
model, optimizer = medzoo.create_model(args)
criterion = DiceLoss(classes=args.classes)
if args.cuda:
torch.cuda.manual_seed(seed)
model = model.cuda()
print("Model transferred in GPU.....")
print("START TRAINING...")
for epoch in range(1, args.nEpochs + 1):
train_stats = train.train_dice(args, epoch, model, training_generator,
optimizer, criterion, train_f, writer)
val_stats = train.test_dice(args, epoch, model, val_generator,
criterion, val_f, writer)
utils.write_train_val_score(writer, epoch, train_stats, val_stats)
model.save_checkpoint(args.save,
epoch,
val_stats[0],
optimizer=optimizer)
# if epoch % 5 == 0:
# utils.visualize_no_overlap(args, full_volume, affine, model, epoch, DIM, writer)
#utils.save_model(model, args, val_stats[0], epoch, best_prec1)
train_f.close()
val_f.close()
def __init__(self, args):
name_model = args.model + "_" + args.dataset_name + "_" + datestr()
self.writer = SummaryWriter(log_dir=args.tb_log_dir,
comment=name_model)
# self.step = 0
# self.mode = ''
self.csv_train, self.csv_val = self.create_stats_files(args.save)
self.dataset_name = args.dataset_name
self.label_names = dict_class_names[args.dataset_name]
self.data = {
"train": dict((label, 0.0) for label in self.label_names),
"val": dict((label, 0.0) for label in self.label_names)
}
self.data['train']['loss'] = 0.0
self.data['val']['loss'] = 0.0
self.data['train']['count'] = 1.0
self.data['val']['count'] = 1.0
self.data['train']['dsc'] = 0.0
self.data['val']['dsc'] = 0.0
def main():
args = get_arguments()
utils.make_dirs(args.save)
train_f, val_f = utils.create_stats_files(args.save)
name_model = args.model + "_" + args.dataset_name + "_" + utils.datestr()
writer = SummaryWriter(log_dir='../runs/' + name_model, comment=name_model)
best_pred = 1.01
samples_train = 200
samples_val = 200
training_generator, val_generator, full_volume, affine = medical_loaders.generate_datasets(
args,
path='.././datasets',
samples_train=samples_train,
samples_val=samples_val)
model, optimizer = medzoo.create_model(args)
criterion = medzoo.DiceLoss2D(args.classes)
if args.cuda:
torch.cuda.manual_seed(seed)
model = model.cuda()
for epoch in range(1, args.nEpochs + 1):
train_stats = train.train_dice(args, epoch, model, training_generator,
optimizer, criterion, train_f, writer)
val_stats = train.test_dice(args, epoch, model, val_generator,
criterion, val_f, writer)
utils.write_train_val_score(writer, epoch, train_stats, val_stats)
best_pred = utils.save_model(model=model,
args=args,
dice_loss=val_stats[0],
epoch=epoch,
best_pred_loss=best_pred)
train_f.close()
val_f.close()
# globals
perfs = []
# actual run
if __name__ == "__main__":
# parse args out
parser = argparse.ArgumentParser()
parser.add_argument("identifier", type=str, help="A uniquely identifying string for this file with format [init_weights_class]_[training_class]")
parser.add_argument("pos_class", type=str, help="Which folder in data/Flickr_2800/ to load for training")
parser.add_argument('-t', '--transplant', type=str, help='The identifier of the transplant to start from')
parser.add_argument('-l', '--lesion_indicator', type=str, default='')
args = parser.parse_args()
# load user configured constants
if args.lesion_indicator is not '':
IDENTIFIER = args.identifier + '_t' + args.lesion_indicator + utils.datestr()
else:
IDENTIFIER = args.identifier + utils.datestr()
POSITIVE_IMAGE_DIR = os.path.join('data/Flickr_2800', args.pos_class)
# let the user know what we're up to
log.log("Training " + IDENTIFIER + " FROM " + POSITIVE_IMAGE_DIR)
if args.transplant is not None:
log.log("(Starting from " + args.transplant + "_t" + args.lesion_indicator + ")")
else:
log.log("(Starting from <blank net, random weights>)")
# train it all
with AlexNet() as alexnet:
# Load data
