XNet: A convolutional neural network (CNN) implementation for medical X-Ray image segmentation suitable for small datasets

Joseph Bullock,Carolina Cuesta-Lazaro,Arnau Quera-Bofarull

DOI: https://doi.org/10.1117/12.2512451

2019-04-20

Abstract:X-Ray image enhancement, along with many other medical image processing applications, requires the segmentation of images into bone, soft tissue, and open beam regions. We apply a machine learning approach to this problem, presenting an end-to-end solution which results in robust and efficient inference. Since medical institutions frequently do not have the resources to process and label the large quantity of X-Ray images usually needed for neural network training, we design an end-to-end solution for small datasets, while achieving state-of-the-art results. Our implementation produces an overall accuracy of 92%, F1 score of 0.92, and an AUC of 0.98, surpassing classical image processing techniques, such as clustering and entropy based methods, while improving upon the output of existing neural networks used for segmentation in non-medical contexts. The code used for this project is available online.

Computer Vision and Pattern Recognition,Artificial Intelligence,Medical Physics

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is medical X - ray image segmentation, especially its application on small - data sets. Specifically, the authors aim to develop a method that can segment X - ray images efficiently and accurately, dividing them into bone, soft - tissue, and open - beam areas. This problem is very important in medical imaging because correct image segmentation can improve the effect of image enhancement, support computer - aided surgery, and contribute to tasks such as anomaly detection. The key challenges of the paper are: 1. **Small - data sets**: Since medical institutions usually lack the resources to process and label a large number of X - ray images, an effective model that can be trained on small - data sets is required. 2. **Noise and boundary continuity**: Traditional image processing techniques (such as clustering methods, entropy - based methods, etc.) perform poorly in dealing with noise and are difficult to generate smooth boundaries. 3. **Generalization ability**: The model needs to be able to generalize well to different body parts, not just specific regions. To address these challenges, the authors designed a unique convolutional neural network (CNN) architecture, named XNet, which can extract fine - grained features while controlling the number of trainable parameters to prevent overfitting. By training on 150 X - ray images and using data augmentation techniques to generate more training samples, XNet achieved a significant performance improvement in the segmentation task, with an overall accuracy of 92%, an F1 - score of 0.92, and an AUC of 0.98. These results outperform traditional image processing techniques and existing neural networks used in non - medical contexts. In addition, the authors also discussed the calibration of the network and how to reduce false positives in the soft - tissue category by setting probability thresholds. Overall, XNet provides a method for efficient and accurate X - ray image segmentation on small - data sets and has important practical application value.

XNet: A convolutional neural network (CNN) implementation for medical X-Ray image segmentation suitable for small datasets

Y-Net for Chest X-Ray Preprocessing: Simultaneous Classification of Geometry and Segmentation of Annotations

Automatic Defect Segmentation in X-Ray Images Based on Deep Learning

X-Net: a dual encoding–decoding method in medical image segmentation

X-Net: Brain Stroke Lesion Segmentation Based on Depthwise Separable Convolution and Long-range Dependencies

EFCNet: Every Feature Counts for Small Medical Object Segmentation

XNet v2: Fewer Limitations, Better Results and Greater Universality

Automated Segmentation of Computed Tomography Images with Submanifold Sparse Convolutional Networks

Machine-Learning-based Algorithms for Automated Image Segmentation Techniques of Transmission X-ray Microscopy (TXM)

N-Net: A novel dense fully convolutional neural network for thyroid nodule segmentation

xLSTM-UNet can be an Effective 2D & 3D Medical Image Segmentation Backbone with Vision-LSTM (ViL) better than its Mamba Counterpart

μ-Net: A Deep Learning-Based Architecture for μ-CT Segmentation

A Hybrid CNN-TransXNet Approach for Advanced Glomerular Segmentation in Renal Histology Imaging

A Novel Architecture for Feature Extraction and Convolution for Image Segmentation of Pathology Detection from Chest X-Ray Images

DS-UNeXt: depthwise separable convolution network with large convolutional kernel for medical image segmentation

nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation

UNet Based Pipeline for Lung Segmentation from Chest X-Ray Images

Neuro-XAI: Explainable deep learning framework based on deeplabV3+ and bayesian optimization for segmentation and classification of brain tumor in MRI scans

nnSegNeXt: A 3D Convolutional Network for Brain Tissue Segmentation Based on Quality Evaluation

High-performance deep learning segmentation for non-destructive testing of X-ray tomography

ConvMedSegNet:A multi-receptive field depthwise convolutional neural network for medical image segmentation