Abstract:Data augmentation is widely applied to medical image analysis tasks in limited datasets with imbalanced classes and insufficient annotations. However, traditional augmentation techniques cannot supply extra information, making the performance of diagnosis unsatisfactory. GAN-based generative methods have thus been proposed to obtain additional useful information to realize more effective data augmentation; but existing generative data augmentation techniques mainly encounter two problems: (i) Current generative data augmentation lacks of the capability in using cross-domain differential information to extend limited datasets. (ii) The existing generative methods cannot provide effective supervised information in medical image segmentation tasks. To solve these problems, we propose an attention-guided cross-domain tumor image generation model (CDA-GAN) with an information enhancement strategy. The CDA-GAN can generate diverse samples to expand the scale of datasets, improving the performance of medical image diagnosis and treatment tasks. In particular, we incorporate channel attention into a CycleGAN-based cross-domain generation network that captures inter-domain information and generates positive or negative samples of brain tumors. In addition, we propose a semi-supervised spatial attention strategy to guide spatial information of features at the pixel level in tumor generation. Furthermore, we add spectral normalization to prevent the discriminator from mode collapse and stabilize the training procedure. Finally, to resolve an inapplicability problem in the segmentation task, we further propose an application strategy of using this data augmentation model to achieve more accurate medical image segmentation with limited data. Experimental studies on two public brain tumor datasets (BraTS and TCIA) show that the proposed CDA-GAN model greatly outperforms the state-of-the-art generative data augmentation in both practical medical image classification tasks and segmentation tasks; e.g. CDA-GAN is 0.50%, 1.72%, 2.05%, and 0.21% better than the best SOTA baseline in terms of ACC, AUC, Recall, and F1, respectively, in the classification task of BraTS, while its improvements w.r.t. the best SOTA baseline in terms of Dice, Sens, HD95, and mIOU, in the segmentation task of TCIA are 2.50%, 0.90%, 14.96%, and 4.18%, respectively.

Red-GAN: Attacking Class Imbalance Via Conditioned Generation. Yet Another Perspective on Medical Image Synthesis for Skin Lesion Dermoscopy and Brain Tumor MRI

An Adversarial Learning Approach to Medical Image Synthesis for Lesion Detection

Medical Image Synthesis for Data Augmentation and Anonymization using Generative Adversarial Networks

Generating Highly Realistic Images of Skin Lesions with GANs

GAN Augmentation: Augmenting Training Data using Generative Adversarial Networks

Red blood cell image generation for data augmentation using Conditional Generative Adversarial Networks

DermGAN: Synthetic Generation of Clinical Skin Images with Pathology

GAN-Based Data Augmentation and Anonymization for Skin-Lesion Analysis: A Critical Review

Synthetic Generation of Dermatoscopic Images with GAN and Closed-Form Factorization

Label-informed cardiac magnetic resonance image synthesis through conditional generative adversarial networks

MelanoGANs: High Resolution Skin Lesion Synthesis with GANs

Cross-domain attention-guided generative data augmentation for medical image analysis with limited data

High-resolution medical image synthesis using progressively grown generative adversarial networks

Medical Image Synthesis with Generative Adversarial Networks for Tissue Recognition

DR-GAN: Conditional Generative Adversarial Network for Fine-Grained Lesion Synthesis on Diabetic Retinopathy Images

A GAN-based image synthesis method for skin lesion classification

Medical image synthesis via conditional GANs: Application to segmenting brain tumours

Generative Adversarial Network for Medical Images (MI-GAN)

SAG-GAN: Semi-Supervised Attention-Guided GANs for Data Augmentation on Medical Images

A Novel Conditional Medical Image Generation GAN with Latent Space Optimal Transport

Enhancing Medical Imaging with GANs Synthesizing Realistic Images from Limited Data