Free-form tumor synthesis in computed tomography images via richer generative adversarial network

TL;DR

This paper introduces a novel 3D tumor synthesis method in CT images using a richer GAN with enhanced features and hybrid loss, improving the realism and diversity of generated tumors for medical imaging applications.

Contribution

The paper presents a new richer convolutional GAN architecture with a hybrid loss for more realistic tumor synthesis in CT images, addressing data scarcity in medical imaging.

Findings

Improved synthesis quality over existing methods.

Effective tumor generation across multiple organ datasets.

Ablation studies confirm the benefits of the richer features and hybrid loss.

Abstract

The insufficiency of annotated medical imaging scans for cancer makes it challenging to train and validate data-hungry deep learning models in precision oncology. We propose a new richer generative adversarial network for free-form 3D tumor/lesion synthesis in computed tomography (CT) images. The network is composed of a new richer convolutional feature enhanced dilated-gated generator (RicherDG) and a hybrid loss function. The RicherDG has dilated-gated convolution layers to enable tumor-painting and to enlarge perceptive fields; and it has a novel richer convolutional feature association branch to recover multi-scale convolutional features especially from uncertain boundaries between tumor and surrounding healthy tissues. The hybrid loss function, which consists of a diverse range of losses, is designed to aggregate complementary information to improve optimization. We perform a comprehensive evaluation of the synthesis results on a wide range of public CT image datasets covering the liver, kidney tumors, and lung nodules. The qualitative and quantitative evaluations and ablation study demonstrated improved synthesizing results over advanced tumor synthesis methods.