Memory-efficient GAN-based Domain Translation of High Resolution 3D Medical Images

TL;DR

This paper introduces a memory-efficient multi-scale patch-based GAN method for high-resolution 3D medical image translation, enabling unpaired domain translation while maintaining image quality and constant GPU memory usage.

Contribution

The proposed multi-scale patch-based GAN approach allows high-resolution 3D medical image translation with fixed memory demand, overcoming computational limitations of traditional GANs.

Findings

Better image quality than patch-based approaches

Prevents patch artifacts in generated images

Enables generation of arbitrarily large images

Abstract

Generative adversarial networks (GANs) are currently rarely applied on 3D medical images of large size, due to their immense computational demand. The present work proposes a multi-scale patch-based GAN approach for establishing unpaired domain translation by generating 3D medical image volumes of high resolution in a memory-efficient way. The key idea to enable memory-efficient image generation is to first generate a low-resolution version of the image followed by the generation of patches of constant sizes but successively growing resolutions. To avoid patch artifacts and incorporate global information, the patch generation is conditioned on patches from previous resolution scales. Those multi-scale GANs are trained to generate realistically looking images from image sketches in order to perform an unpaired domain translation. This allows to preserve the topology of the test data and generate the appearance of the training domain data. The evaluation of the domain translation scenarios is performed on brain MRIs of size 155x240x240 and thorax CTs of size up to 512x512x512. Compared to common patch-based approaches, the multi-resolution scheme enables better image quality and prevents patch artifacts. Also, it ensures constant GPU memory demand independent from the image size, allowing for the generation of arbitrarily large images.