English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Unpaired Medical Image Enhancement (UMIE) aims to transform a low-quality(LQ) medical image into a high-quality (HQ) one without relying on pairedimages for training. While most existing approaches are based onPix2Pix/CycleGAN and are effective to some extent, they fail to explicitly useHQ information to guide the enhancement process, which can lead to undesiredartifacts and structural distortions. In this paper, we propose a novel UMIEapproach that avoids the above limitation of existing methods by directlyencoding HQ cues into the LQ enhancement process in a variational fashion andthus model the UMIE task under the joint distribution between the LQ and HQdomains. Specifically, we extract features from an HQ image and explicitlyinsert the features, which are expected to encode HQ cues, into the enhancementnetwork to guide the LQ enhancement with the variational normalization module.We train the enhancement network adversarially with a discriminator to ensurethe generated HQ image falls into the HQ domain. We further propose acontent-aware loss to guide the enhancement process with wavelet-basedpixel-level and multi-encoder-based feature-level constraints. Additionally, asa key motivation for performing image enhancement is to make the enhancedimages serve better for downstream tasks, we propose a bi-level learning schemeto optimize the UMIE task and downstream tasks cooperatively, helping generateHQ images both visually appealing and favorable for downstream tasks.Experiments on three medical datasets, including two newly collected datasets,verify that the proposed method outperforms existing techniques in terms ofboth enhancement quality and downstream task performance. We will make the codeand the newly collected datasets publicly available for community study.

HQG-Net: Unpaired Medical Image Enhancement with High-Quality Guidance