Abstract:Background: Deep neural networks have been widely used in medical image segmentation and have achieved state-of-the-art performance in many tasks. However, different from the segmentation of natural images or video frames, the manual segmentation of anatomical structures in medical images needs high expertise so the scale of labeled training data is very small, which is a major obstacle for the improvement of deep neural networks performance in medical image segmentation. Methods: In this paper, we proposed a new end-to-end generation-segmentation framework by integrating Generative Adversarial Networks (GAN) and a segmentation network and train them simultaneously. The novelty is that during the training of the GAN, the intermediate synthetic images generated by the generator of the GAN are used to pre-train the segmentation network. As the advances of the training of the GAN, the synthetic images evolve gradually from being very coarse to containing more realistic textures, and these images help train the segmentation network gradually. After the training of GAN, the segmentation network is then fine-tuned by training with the real labeled images. Results: We evaluated the proposed framework on four different datasets, including 2D cardiac dataset and lung dataset, 3D prostate dataset and liver dataset. Compared with original U-net and CE-Net, our framework can achieve better segmentation performance. Our framework also can get better segmentation results than U-net on small datasets. In addition, our framework is more effective than the usual data augmentation methods. Conclusions: The proposed framework can be used as a pre-train method of segmentation network, which helps to get a better segmentation result. Our method can solve the shortcomings of current data augmentation methods to some extent.

UENet: A Novel Generative Adversarial Network for Angiography Image Segmentation.

Generative Adversarial Network Based Cerebrovascular Segmentation for Time-of-flight Magnetic Resonance Angiography Image

DeU-Net 2.0: Enhanced deformable U-Net for 3D cardiac cine MRI segmentation

CAS‐GAN: A Novel Generative Adversarial Network‐Based Architecture for Coronary Artery Segmentation

Optimizing ensemble U-Net architectures for robust coronary vessel segmentation in angiographic images

Retinal Vessel Segmentation Combined with Generative Adversarial Networks and Dense U-Net

Centerline-supervision multi-task learning network for coronary angiography segmentation

Graph Convolution Based Cross-Network Multiscale Feature Fusion for Deep Vessel Segmentation

UGAN: Semi-supervised Medical Image Segmentation Using Generative Adversarial Network

CAU-net: A Novel Convolutional Neural Network for Coronary Artery Segmentation in Digital Substraction Angiography.

The Domain Shift Problem of Medical Image Segmentation and Vendor-Adaptation by Unet-GAN

End-to-End Automatic Classification of Retinal Vessel Based on Generative Adversarial Networks with Improved U-Net.

GANet: Group Attention Network for Diabetic Retinopathy Image Segmentation

A Generative Adversarial Framework for Capillary Non-perfusion Regions Segmentation in Fundus Fluorescein Angiograms

Generative Adversarial Training with Dual-Attention for Vascular Segmentation and Topological Analysis

Annotation-Free Cardiac Vessel Segmentation via Knowledge Transfer from Retinal Images

Genetic U-Net: Automatically Designed Deep Networks for Retinal Vessel Segmentation Using a Genetic Algorithm

Multichannel Fully Convolutional Network for Coronary Artery Segmentation in X-Ray Angiograms

A Refined Equilibrium Generative Adversarial Network for Retinal Vessel Segmentation

Automated Coronary Tree Segmentation for X-ray Angiography Sequences Using Fully-convolutional Neural Networks

Generative Adversarial Networks for Pre-training of Medical Image Segmentation Networks