AdaIN-Switchable CycleGAN for Efficient Unsupervised Low-Dose CT Denoising

TL;DR

This paper introduces an efficient unsupervised low-dose CT denoising method using a novel AdaIN-Switchable CycleGAN with a single generator, reducing memory and parameter requirements while improving performance.

Contribution

A new cycleGAN architecture with a switchable generator based on AdaIN layers that minimizes memory use and simplifies training for low-dose CT denoising.

Findings

Outperforms previous cycleGAN methods in quality.

Uses approximately half the parameters of traditional models.

Enables more stable training with limited data.

Abstract

Recently, deep learning approaches have been extensively studied for low-dose CT denoising thanks to its superior performance despite the fast computational time. In particular, cycleGAN has been demonstrated as a powerful unsupervised learning scheme to improve the low-dose CT image quality without requiring matched high-dose reference data. Unfortunately, one of the main limitations of the cycleGAN approach is that it requires two deep neural network generators at the training phase, although only one of them is used at the inference phase. The secondary auxiliary generator is needed to enforce the cycle-consistency, but the additional memory requirement and increases of the learnable parameters are the main huddles for cycleGAN training. To address this issue, here we propose a novel cycleGAN architecture using a single switchable generator. In particular, a single generator is implemented using adaptive instance normalization (AdaIN) layers so that the baseline generator converting a low-dose CT image to a routine-dose CT image can be switched to a generator converting high-dose to low-dose by simply changing the AdaIN code. Thanks to the shared baseline network, the additional memory requirement and weight increases are minimized, and the training can be done more stably even with small training data. Experimental results show that the proposed method outperforms the previous cycleGAN approaches while using only about half the parameters.