Metadata-Version: 2.1
Name: torchio
Version: 0.7.5
Summary: Tools for loading, augmenting and writing 3D medical images on PyTorch.
Home-page: https://github.com/fepegar/torchio
Author: Fernando Perez-Garcia
Author-email: fernando.perezgarcia.17@ucl.ac.uk
License: MIT license
Description: # TorchIO
        
        [![PyPI version](https://badge.fury.io/py/torchio.svg)](https://badge.fury.io/py/torchio)
        [![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.3598623.svg)](https://doi.org/10.5281/zenodo.3598623)
        
        
        `torchio` is a Python package containing a set of tools to efficiently
        read, sample and write 3D medical images in deep learning applications
        written in [PyTorch](https://pytorch.org/),
        including intensity and spatial transforms
        for data augmentation and preprocessing. Transforms include typical computer vision operations
        such as random affine transformations and also domain specific ones such as
        simulation of intensity artifacts due to
        [MRI magnetic field inhomogeneity](http://mriquestions.com/why-homogeneity.html)
        or [k-space motion artifacts](http://proceedings.mlr.press/v102/shaw19a.html).
        
        This package has been greatly inspired by [NiftyNet](https://niftynet.io/).
        
        
        ## Index
        
        - [Installation](#installation)
        - [Features](#features)
          * [Data handling](#data-handling)
            - [`ImagesDataset`](#imagesdataset)
            - [Samplers](#samplers)
            - [`Queue`](#queue)
          * [Transforms](#transforms)
            - [Intensity](#intensity)
              * [MRI k-space motion artifacts](#mri-k-space-motion-artifacts)
              * [MRI magnetic field inhomogeneity](#mri-magnetic-field-inhomogeneity)
              * [Gaussian noise](#gaussian-noise)
              * [Normalization](#normalization)
                - [Histogram standardization](#histogram-standardization)
                - [Z-normalization](#z-normalization)
                - [Rescale](#rescale)
            - [Spatial](#spatial)
              * [Flip](#flip)
              * [Affine transform](#affine-transform)
              * [B-spline dense deformation](#b-spline-dense-deformation)
        - [Example](#example)
        
        
        ## Installation
        
        ```shell
        $ pip install torchio
        ```
        
        
        ## Features
        
        ### Data handling
        
        #### [`ImagesDataset`](torchio/dataset/images.py)
        
        `ImagesDataset` is a reader of medical images that directly inherits from
        [`torch.utils.Dataset`](https://pytorch.org/docs/stable/data.html#torch.utils.data.Dataset).
        It can be used with a
        [`torch.utils.DataLoader`](https://pytorch.org/docs/stable/data.html#torch.utils.data.DataLoader)
        for efficient reading and data augmentation.
        
        The paths suffix must be `.nii`, `.nii.gz` or `.nrrd`.
        
        ```python
        import torchio
        
        subject_a = {
            't1': dict(path='~/Dropbox/MRI/t1.nii.gz', type=torchio.INTENSITY),
            'label': dict(path='~/Dropbox/MRI/t1_seg.nii.gz', type=torchio.LABEL),
        }
        subject_b = {
            't1': dict(path='/tmp/colin27_t1_tal_lin.nii.gz', type=torchio.INTENSITY),
            'label': dict(path='/tmp/colin27_seg1.nii.gz', type=torchio.LABEL),
        }
        paths_list = [subject_a, subject_b]
        subjects_dataset = torchio.ImagesDataset(paths_list)
        subject_sample = subjects_dataset[0]
        ```
        
        
        #### [Samplers](torchio/sampler/sampler.py)
        
        `torchio` includes grid, uniform and label patch samplers. There is also an
        aggregator used for dense predictions. The code for these is almost
        [copy-pasted from NiftyNet](https://github.com/NifTK/NiftyNet/tree/dev/niftynet/engine).
        
        For more information about patch-based training, see
        [NiftyNet docs](https://niftynet.readthedocs.io/en/dev/window_sizes.html).
        
        ```python
        import torch
        import torchio
        
        CHANNELS_DIMENSION = 1
        patch_overlap = 4
        grid_sampler = torchio.inference.GridSampler(
            input_array,  # some NumPy array
            patch_size=128,
            patch_overlap=patch_overlap,
        )
        patch_loader = torch.utils.data.DataLoader(grid_sampler, batch_size=4)
        aggregator = torchio.inference.GridAggregator(
            input_array,
            patch_overlap=patch_overlap,
        )
        
        with torch.no_grad():
            for patches_batch in patch_loader:
                input_tensor = patches_batch['image']
                locations = patches_batch['location']
                logits = model(input_tensor)  # some torch.nn.Module
                labels = logits.argmax(dim=CHANNELS_DIMENSION, keepdim=True)
                outputs = labels
                aggregator.add_batch(outputs, locations)
        
        output_array = aggregator.output_array
        ```
        
        
        #### [`Queue`](torchio/queue.py)
        
        A patches `Queue` (or buffer) can be used for randomized patch-based sampling
        during training.
        [This interactive animation](https://niftynet.readthedocs.io/en/dev/config_spec.html#queue-length)
        can be used to understand how the queue works.
        
        ```python
        import torch
        import torchio
        
        patches_queue = torchio.Queue(
            subjects_dataset=subjects_dataset,
            queue_length=300,
            samples_per_volume=10,
            patch_size=96,
            sampler_class=torchio.sampler.ImageSampler,
            num_workers=4,
            shuffle_subjects=True,
            shuffle_patches=True,
        )
        patches_loader = DataLoader(patches_queue, batch_size=4)
        
        num_epochs = 20
        for epoch_index in range(num_epochs):
            for patches_batch in patches_loader:
                logits = model(patches_batch)  # model is some torch.nn.Module
        ```
        
        
        ### Transforms
        
        The transforms package should remind users of
        [`torchvision.transforms`](https://pytorch.org/docs/stable/torchvision/transforms.html).
        They take as input the samples generated by an
        [`ImagesDataset`](#dataset).
        
        #### Intensity
        
        ##### [MRI k-space motion artifacts](torchio/transforms/random_motion.py)
        
        Magnetic resonance images suffer from motion artifacts when the subject moves
        during image acquisition. This transform follows
        [Shaw et al., 2019](http://proceedings.mlr.press/v102/shaw19a.html) to
        simulate motion artifacts for data augmentation.
        
        ![MRI k-space motion artifacts](images/random_motion.gif)
        
        
        ##### [MRI magnetic field inhomogeneity](torchio/transforms/random_bias_field.py)
        
        MRI magnetic field inhomogeneity creates slow frequency intensity variations.
        This transform is very similar to the one in
        [NiftyNet](https://niftynet.readthedocs.io/en/dev/niftynet.layer.rand_bias_field.html).
        
        ![MRI bias field artifacts](images/random_bias_field.gif)
        
        
        ##### [Gaussian noise](torchio/transforms/random_noise.py)
        
        Adds noise sampled from a normal distribution with mean 0 and standard
        deviation sampled from a uniform distribution in the range `std_range`.
        It is often used after [`ZNormalization`](#z-normalization), as the output of
        this transform has zero-mean.
        
        ![Random Gaussian noise](images/random_noise.gif)
        
        
        ##### Normalization
        
        ###### [Histogram standardization](torchio/transforms/histogram_standardization.py)
        
        Implementation of
        [*New variants of a method of MRI scale standardization*](https://ieeexplore.ieee.org/document/836373)
        adapted from NiftyNet.
        
        ![Histogram standardization](images/histogram_standardization.png)
        
        
        ###### [Z-normalization](torchio/transforms/z_normalization.py)
        
        
        ###### [Rescale](torchio/transforms/rescale.py)
        
        
        #### Spatial
        
        ##### [Flip](torchio/transforms/random_flip.py)
        
        Reverse the order of elements in an image along the given axes.
        
        
        ##### [Affine transform](torchio/transforms/random_affine.py)
        
        
        ##### [B-spline dense elastic deformation](torchio/transforms/random_elastic_deformation.py)
        
        ![Random elastic deformation](images/random_elastic_deformation.gif)
        
        
        ## [Example](examples/example_times.py)
        
        This example shows the improvement in performance when multiple workers are
        used to load and preprocess the volumes using multiple workers.
        
        ```python
        import time
        import multiprocessing as mp
        
        from torch.utils.data import DataLoader
        from torchvision.transforms import Compose
        
        from torchio import ImagesDataset, Queue
        from torchio.sampler import ImageSampler
        from torchio.utils import create_dummy_dataset
        from torchio.transforms import (
            ZNormalization,
            RandomNoise,
            RandomFlip,
            RandomAffine,
        )
        
        
        # Define training and patches sampling parameters
        num_epochs = 4
        patch_size = 128
        queue_length = 100
        samples_per_volume = 10
        batch_size = 4
        
        def model(batch, sleep_time=0.1):
            """Dummy function to simulate a forward pass through the network"""
            time.sleep(sleep_time)
            return batch
        
        # Create a dummy dataset in the temporary directory, for this example
        subjects_paths = create_dummy_dataset(
            num_images=100,
            size_range=(193, 229),
            force=False,
        )
        
        # Each element of subjects_paths is a dictionary:
        # subject = {
        #     'one_image': dict(path=path_to_one_image, type=torchio.INTENSITY),
        #     'another_image': dict(path=path_to_another_image, type=torchio.INTENSITY),
        #     'a_label': dict(path=path_to_a_label, type=torchio.LABEL),
        # }
        
        # Define transforms for data normalization and augmentation
        transforms = (
            ZNormalization(),
            RandomNoise(std_range=(0, 0.25)),
            RandomAffine(scales=(0.9, 1.1), degrees=10),
            RandomFlip(axes=(0,)),
        )
        transform = Compose(transforms)
        subjects_dataset = ImagesDataset(subjects_paths, transform)
        
        sample = subjects_dataset[0]
        
        # Run a benchmark for different numbers of workers
        workers = range(mp.cpu_count() + 1)
        for num_workers in workers:
            print('Number of workers:', num_workers)
        
            # Define the dataset as a queue of patches
            queue_dataset = Queue(
                subjects_dataset,
                queue_length,
                samples_per_volume,
                patch_size,
                ImageSampler,
                num_workers=num_workers,
            )
            batch_loader = DataLoader(queue_dataset, batch_size=batch_size)
        
            start = time.time()
            for epoch_index in range(num_epochs):
                for batch in batch_loader:
                    logits = model(batch)
            print('Time:', int(time.time() - start), 'seconds')
            print()
        ```
        
        
        Output:
        ```python
        Number of workers: 0
        Time: 394 seconds
        
        Number of workers: 1
        Time: 372 seconds
        
        Number of workers: 2
        Time: 278 seconds
        
        Number of workers: 3
        Time: 259 seconds
        
        Number of workers: 4
        Time: 242 seconds
        ```
        
        
        =======
        History
        =======
        
        0.2.0 (2019-12-06)
        ------------------
        
        * First release on PyPI.
        
        
        0.3.0 (21-12-2019)
        ------------------
        
        * Add Rescale transform
        * Add support for multimodal data and missing modalities
        
        
        0.4.0 (29-12-2019)
        ------------------
        
        * Add MRI k-space motion artefact augmentation
        
        
        0.5.0 (01-01-2020)
        ------------------
        
        * Add bias field transform
        
        
        0.6.0 (02-01-2020)
        ------------------
        
        * Add support for NRRD
        
        
        0.7.0 (02-01-2020)
        ------------------
        
        * Make transforms use PyTorch tensors consistently
        
Keywords: torchio
Platform: UNKNOWN
Classifier: Development Status :: 2 - Pre-Alpha
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: MIT License
Classifier: Natural Language :: English
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.6
Classifier: Programming Language :: Python :: 3.7
Requires-Python: >=3.6
Description-Content-Type: text/markdown
