SupervisedTrainer and non-binary label

[Gijz33]

Hi!

How does supervisedtrainer handle non-binary labels (between 0 and 1) when a binary label is requested.
Does the model assume labelpixel >0 --> 1. Or does it something like: everything <0.5 --> 0 and >0.5 --> 1.

[Nic-Ma]

Hi @Gijz33 ,

Thanks for your interest here.
I think your question is how to apply the postprocessing transforms for label in SupervisedTrainer.
You can refer to this example:
https://github.com/Project-MONAI/tutorials/blob/master/modules/engines/unet_evaluation_dict.py#L90

Thanks.

[Gijz33]

Hi @Nic-Ma,
I see my question was not clearly stated. I'm sorry for that. I'll try again:)

I have trained a model using training data with non-binary labels. The pixel values of these labels are between 0 and 1.
I expected an error, because my model asks for binary labels. However, the training has been completed.
I was wondering how the model did interpret the non-binary labels from the training data. Does is use a threshold value to determine if a pixel is positive or negative?

[Nic-Ma]

Hi @Gijz33 ,

Could you please help share your test program for further analysis?

Thanks.

[Gijz33]

Hi @Nic-Ma,

I'm sorry for my late response.
This is the model I'm using:

# Copyright 2020 MONAI Consortium
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#     http://www.apache.org/licenses/LICENSE-2.0
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import argparse
import glob
import logging
import os
import shutil
import sys

import numpy as np
import torch
import torch.nn as nn
from ignite.contrib.handlers import ProgressBar

import monai
from monai.handlers import CheckpointSaver, MeanDice, StatsHandler, ValidationHandler, from_engine
from monai.transforms import (
    AddChanneld,
    AsDiscreted,
    CastToTyped,
    LoadImaged,
    Orientationd,
    RandAffined,
    RandCropByPosNegLabeld,
    RandFlipd,
    RandGaussianNoised,
    ScaleIntensityRanged,
    Spacingd,
    SpatialPadd,
    EnsureTyped,
)


def get_xforms(mode="train", keys=("image", "label")):
    """returns a composed transform for train/val/infer."""

    xforms = [
        LoadImaged(keys),
        AddChanneld(keys),
        Orientationd(keys, axcodes="LPS"),
        Spacingd(keys, pixdim=(1.25, 1.25, 5.0), mode=("bilinear", "nearest")[: len(keys)]),
        ScaleIntensityRanged(keys[0], a_min=-1000.0, a_max=500.0, b_min=0.0, b_max=1.0, clip=True),
    ]
    if mode == "train":
        xforms.extend(
            [
                SpatialPadd(keys, spatial_size=(192, 192, -1), mode="reflect"),  # ensure at least 192x192
                RandAffined(
                    keys,
                    prob=0.15,
                    rotate_range=(0.05, 0.05, None),  # 3 parameters control the transform on 3 dimensions
                    scale_range=(0.1, 0.1, None),
                    mode=("bilinear", "nearest"),
                    as_tensor_output=False,
                ),
                RandCropByPosNegLabeld(keys, label_key=keys[1], spatial_size=(192, 192, 16), num_samples=3),
                RandGaussianNoised(keys[0], prob=0.15, std=0.01),
                RandFlipd(keys, spatial_axis=0, prob=0.5),
                RandFlipd(keys, spatial_axis=1, prob=0.5),
                RandFlipd(keys, spatial_axis=2, prob=0.5),
            ]
        )
        dtype = (np.float32, np.uint8)
    if mode == "val":
        dtype = (np.float32, np.uint8)
    if mode == "infer":
        dtype = (np.float32,)
    xforms.extend([CastToTyped(keys, dtype=dtype), EnsureTyped(keys)])
    return monai.transforms.Compose(xforms)


def get_net():
    """returns a unet model instance."""
    n_classes=2
    net=monai.networks.nets.UNet(
        dimensions=3,
        in_channels=1,
        out_channels=n_classes,
        channels=(16, 32, 64, 128, 256),
        strides=(2, 2, 2, 2),
        #num_res_units=2,
        dropout=0.5,
    )
    return net


def get_inferer(_mode=None):
    """returns a sliding window inference instance."""

    patch_size = (192, 192, 16)
    sw_batch_size, overlap = 2, 0.5
    inferer = monai.inferers.SlidingWindowInferer(
        roi_size=patch_size,
        sw_batch_size=sw_batch_size,
        overlap=overlap,
        mode="gaussian",
        padding_mode="replicate",
    )
    return inferer


class DiceCELoss(nn.Module):
    """Dice and Xentropy loss"""

    def __init__(self):
        super().__init__()
        self.dice = monai.losses.DiceLoss(to_onehot_y=True, softmax=True)
        self.cross_entropy = nn.CrossEntropyLoss()

    def forward(self, y_pred, y_true):
        dice = self.dice(y_pred, y_true)
        # CrossEntropyLoss target needs to have shape (B, D, H, W)
        # Target from pipeline has shape (B, 1, D, H, W)
        cross_entropy = self.cross_entropy(y_pred, torch.squeeze(y_true, dim=1).long())
        return dice + cross_entropy


def train(data_folder=".", model_folder="runs"):
    """run a training pipeline."""

    images = sorted(glob.glob(os.path.join(data_folder, "*_ct.nii.gz")))
    labels = sorted(glob.glob(os.path.join(data_folder, "*_seg.nii.gz")))
    logging.info(f"training: image/label ({len(images)}) folder: {data_folder}")

    amp = True  # auto. mixed precision
    keys = ("image", "label")
    train_frac, val_frac = 0.8, 0.2
    n_train = int(train_frac * len(images)) + 1
    n_val = min(len(images) - n_train, int(val_frac * len(images)))
    logging.info(f"training: train {n_train} val {n_val}, folder: {data_folder}")

    train_files = [{keys[0]: img, keys[1]: seg} for img, seg in zip(images[:n_train], labels[:n_train])]
    val_files = [{keys[0]: img, keys[1]: seg} for img, seg in zip(images[-n_val:], labels[-n_val:])]

    # create a training data loader
    batch_size = 2
    logging.info(f"batch size {batch_size}")
    train_transforms = get_xforms("train", keys)
    train_ds = monai.data.CacheDataset(data=train_files, transform=train_transforms)
    train_loader = monai.data.DataLoader(
        train_ds,
        batch_size=batch_size,
        shuffle=True,
        num_workers=2,
        pin_memory=torch.cuda.is_available(),
    )

    # create a validation data loader
    val_transforms = get_xforms("val", keys)
    val_ds = monai.data.CacheDataset(data=val_files, transform=val_transforms)
    val_loader = monai.data.DataLoader(
        val_ds,
        batch_size=1,  # image-level batch to the sliding window method, not the window-level batch
        num_workers=2,
        pin_memory=torch.cuda.is_available(),
    )

    # create BasicUNet, DiceLoss and Adam optimizer
    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
    net = get_net().to(device)
    max_epochs, lr, momentum = 500, 1e-4, 0.95
    logging.info(f"epochs {max_epochs}, lr {lr}, momentum {momentum}")
    opt = torch.optim.Adam(net.parameters(), lr=lr)

    # create evaluator (to be used to measure model quality during training
    val_post_transform = monai.transforms.Compose(
        [EnsureTyped(keys=("pred", "label")), AsDiscreted(keys=("pred", "label"), argmax=(True, False), to_onehot=2)]
    )
    val_handlers = [
        ProgressBar(),
        CheckpointSaver(save_dir=model_folder, save_dict={"net": net}, save_key_metric=True, key_metric_n_saved=3),
    ]
    evaluator = monai.engines.SupervisedEvaluator(
        device=device,
        val_data_loader=val_loader,
        network=net,
        inferer=get_inferer(),
        postprocessing=val_post_transform,
        key_val_metric={
            "val_mean_dice": MeanDice(include_background=False, output_transform=from_engine(["pred", "label"]))
        },
        val_handlers=val_handlers,
        amp=amp,
    )

    # evaluator as an event handler of the trainer
    train_handlers = [
        ValidationHandler(validator=evaluator, interval=1, epoch_level=True),
        StatsHandler(tag_name="train_loss", output_transform=from_engine(["loss"], first=True)),
    ]
    trainer = monai.engines.SupervisedTrainer(
        device=device,
        max_epochs=max_epochs,
        train_data_loader=train_loader,
        network=net,
        optimizer=opt,
        loss_function=DiceCELoss(),
        inferer=get_inferer(),
        key_train_metric=None,
        train_handlers=train_handlers,
        amp=amp,
    )
    trainer.run()


def infer(data_folder=".", model_folder="runs", prediction_folder="output"):
    """
    run inference, the output folder will be "./output"
    """
    ckpts = sorted(glob.glob(os.path.join(model_folder, "*.pt")))
    ckpt = ckpts[-1]
    for x in ckpts:
        logging.info(f"available model file: {x}.")
    logging.info("----")
    logging.info(f"using {ckpt}.")

    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
    net = get_net().to(device)
    net.load_state_dict(torch.load(ckpt, map_location=device))
    net.eval()

    image_folder = os.path.abspath(data_folder)
    images = sorted(glob.glob(os.path.join(image_folder, "*_ct.nii.gz")))
    logging.info(f"infer: image ({len(images)}) folder: {data_folder}")
    infer_files = [{"image": img} for img in images]

    keys = ("image",)
    infer_transforms = get_xforms("infer", keys)
    infer_ds = monai.data.Dataset(data=infer_files, transform=infer_transforms)
    infer_loader = monai.data.DataLoader(
        infer_ds,
        batch_size=1,  # image-level batch to the sliding window method, not the window-level batch
        num_workers=2,
        pin_memory=torch.cuda.is_available(),
    )

    inferer = get_inferer()
    saver = monai.data.NiftiSaver(output_dir=prediction_folder, mode="nearest")
    with torch.no_grad():
        for infer_data in infer_loader:
            logging.info(f"segmenting {infer_data['image_meta_dict']['filename_or_obj']}")
            preds = inferer(infer_data[keys[0]].to(device), net)
            n = 1.0
            for _ in range(4):
                # test time augmentations
                _img = RandGaussianNoised(keys[0], prob=1.0, std=0.01)(infer_data)[keys[0]]
                pred = inferer(_img.to(device), net)
                preds = preds + pred
                n = n + 1.0
                for dims in [[2], [3]]:
                    flip_pred = inferer(torch.flip(_img.to(device), dims=dims), net)
                    pred = torch.flip(flip_pred, dims=dims)
                    preds = preds + pred
                    n = n + 1.0
            preds = preds / n
            preds = (preds.argmax(dim=1, keepdims=True)).float()
            saver.save_batch(preds, infer_data["image_meta_dict"])

    # copy the saved segmentations into the required folder structure for submission
    submission_dir = os.path.join(prediction_folder, "to_submit")
    if not os.path.exists(submission_dir):
        os.makedirs(submission_dir)
    files = glob.glob(os.path.join(prediction_folder, "volume*", "*.nii.gz"))
    for f in files:
        new_name = os.path.basename(f)
        new_name = new_name[len("volume-covid19-A-0"):]
        new_name = new_name[: -len("_ct_seg.nii.gz")] + ".nii.gz"
        to_name = os.path.join(submission_dir, new_name)
        shutil.copy(f, to_name)
    logging.info(f"predictions copied to {submission_dir}.")


if __name__ == "__main__":
    """
    Usage:
        python run_net.py train --data_folder "COVID-19-20_v2/Train" # run the training pipeline
        python run_net.py infer --data_folder "COVID-19-20_v2/Validation" # run the inference pipeline
    """
    parser = argparse.ArgumentParser(description="Run a basic UNet segmentation baseline.")
    parser.add_argument(
        "mode", metavar="mode", default="train", choices=("train", "infer"), type=str, help="mode of workflow"
    )
    parser.add_argument("--data_folder", default="", type=str, help="training data folder")
    parser.add_argument("--model_folder", default="runs", type=str, help="model folder")
    args = parser.parse_args()

    monai.config.print_config()
    monai.utils.set_determinism(seed=0)
    logging.basicConfig(stream=sys.stdout, level=logging.INFO)

    if args.mode == "train":
        data_folder = args.data_folder or os.path.join("COVID-19-20_v2", "Train")
        train(data_folder=data_folder, model_folder=args.model_folder)
    elif args.mode == "infer":
        data_folder = args.data_folder or os.path.join("COVID-19-20_v2", "Validation")
        infer(data_folder=data_folder, model_folder=args.model_folder)
    else:
        raise ValueError("Unknown mode.")

[Nic-Ma]

Hi @Gijz33 ,

I think self.dice = monai.losses.DiceLoss(to_onehot_y=True, softmax=True) and AsDiscreted(keys=("pred", "label"), argmax=(True, False), to_onehot=2) in your code will convert label to binary values when computing loss or metrics.

Thanks.

[Gijz33]

Hi @Nic-Ma,
Thank you for your answer. I didn't realize that, thanks!

[Gijz33]

Hi @Nic-Ma ,

Regarding the SupervisedEvaluator in the model I sent here, I have got a question about its function.
Does it only evaluate the results of the model being trained, or does it also fine-tune the parameters of the model?

Best,
Gijs

[Nic-Ma]

Only evaluate the model.

Thanks.

[Gijz33]

Thank you for your answer!