# Continuous Conversion of CT Kernel using Switchable CycleGAN with AdaIN

arXiv: https://arxiv.org/abs/2011.13150 · HTML: https://mlsanity.com/paper/2011.13150/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 26 Nov 2020.

## Built on by (most cited first, 8 of 8)

- [Unsupervised Deep Learning Methods for Biological Image Reconstruction and Enhancement](https://mlsanity.com/paper/2105.08040/) (2021)
- [DC-cycleGAN: Bidirectional CT-to-MR Synthesis from Unpaired Data](https://mlsanity.com/paper/2211.01293/) (2022)
- [Federated CycleGAN for Privacy-Preserving Image-to-Image Translation](https://mlsanity.com/paper/2106.09246/) (2021)
- [Feature Disentanglement in generating three-dimensional structure from two-dimensional slice with sliceGAN](https://mlsanity.com/paper/2105.00194/) (2021)
- [Texture-Aware StarGAN for CT data harmonisation](https://mlsanity.com/paper/2503.15058/) (2025)
- [Tunable Image Quality Control of 3-D Ultrasound using Switchable CycleGAN](https://mlsanity.com/paper/2112.02896/) (2021)
- [Multipath cycleGAN for harmonization of paired and unpaired low-dose lung computed tomography reconstruction kernels](https://mlsanity.com/paper/2505.22568/) (2025)
- [One CT Unified Model Training Framework to Rule All Scanning Protocols](https://mlsanity.com/paper/2603.15025/) (2026)

## Abstract (arXiv)

X-ray computed tomography (CT) uses different filter kernels to highlight different structures. Since the raw sinogram data is usually removed after the reconstruction, in case there are additional need for other types of kernel images that were not previously generated, the patient may need to be scanned again. Accordingly, there exists increasing demand for post-hoc image domain conversion from one kernel to another without sacrificing the image quality. In this paper, we propose a novel unsupervised continuous kernel conversion method using cycle-consistent generative adversarial network (cycleGAN) with adaptive instance normalization (AdaIN). Even without paired training data, not only can our network translate the images between two different kernels, but it can also convert images along the interpolation path between the two kernel domains. We also show that the quality of generated images can be further improved if intermediate kernel domain images are available. Experimental results confirm that our method not only enables accurate kernel conversion that is comparable to supervised learning methods, but also generates intermediate kernel images in the unseen domain that are useful for hypopharyngeal cancer diagnosis.