XNet

XNet is a Convolutional Neural Network designed for the segmentation of X-Ray images into bone, soft tissue and open beam regions. Specifically, it performs well on small datasets with the aim to minimise the number of false positives in the soft tissue class.

This code accompanies the paper published in the SPIE Medical Imaging Conference Proceedings (2019) and can be found on the preprint arXiv at: arXiv:1812.00548

Cite as:

@inproceedings{10.1117/12.2512451,
author = {Joseph Bullock and Carolina Cuesta-Lázaro and Arnau Quera-Bofarull},
title = {{XNet: a convolutional neural network (CNN) implementation for medical x-ray image segmentation suitable for small datasets}},
volume = {10953},
booktitle = {Medical Imaging 2019: Biomedical Applications in Molecular, Structural, and Functional Imaging},
editor = {Barjor Gimi and Andrzej Krol},
organization = {International Society for Optics and Photonics},
publisher = {SPIE},
pages = {453 -- 463},
keywords = {machine learning, deep learning, X-Ray segmentation, neural network, small datasets},
year = {2019},
doi = {10.1117/12.2512451},
URL = {https://doi.org/10.1117/12.2512451}
}

Architecture

Built on a typical encoder-decoder architecture as inspired by SegNet.
Additional feature extraction stage, with weight sharing across some layers.
Fine and coarse grained feature preservation through concatenation of layers.
L2 regularisation at each of the convolutional layers, to decrease overfitting.

The architecture is described in the XNet.py file.

Output

XNet outputs a mask of equal size to the input images.

Training

To train a model:

Open Training/generate_parameters.py and define your desired hyperparameters
Run Training/generate_parameters.py to generate a paramteres.txt file which is read Training/TrainingClass.py
Run train.py

XNet is trained on a small dataset which has undergone augmention. Examples of this augmentation step can be found in the Augmentations/augmentations.ipynb notebook. Similarly the Training folder contains python scripts that perform the necessary augementations.

Running Training/train.py calls various other scripts to perform one of two possible ways of augmenting the images:

'On the fly augmentation' where a new set of augmentations is generated at each epoch.
Pre-augmented images.

To select which method to use comment out the corresponding lines in the fit function in the Training/TrainingClass.py script.

train.py also performs postprocessing to fine tune the results.

Benchmarking

XNet was benchmarked against two of the leading segmentation networks:

Simplified SegNet (found in the SimpleSegNet.py file)
UNet (found in the UNet.py file)

Data

We trained on a dataset of:

150 X-Ray images.
No scatter correction.
1500x1500 .tif image downsampled to 200x200
20 human body part classes.
Highly imbalanced.

As this work grew out of work with a corporation we are sadly unable to share the propriatory data we used.

More information

For more information and context see the conference poster Poster.pdf.

Please note that some of the path variables may need to be corrected in order to utilise the current filing system. These are planned to be updated in the future.