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

This paper introduces an innovative methodology for producing high-quality 3Dlung CT images guided by textual information. While diffusion-based generativemodels are increasingly used in medical imaging, current state-of-the-artapproaches are limited to low-resolution outputs and underutilize radiologyreports' abundant information. The radiology reports can enhance the generationprocess by providing additional guidance and offering fine-grained control overthe synthesis of images. Nevertheless, expanding text-guided generation tohigh-resolution 3D images poses significant memory and anatomicaldetail-preserving challenges. Addressing the memory issue, we introduce ahierarchical scheme that uses a modified UNet architecture. We start bysynthesizing low-resolution images conditioned on the text, serving as afoundation for subsequent generators for complete volumetric data. To ensurethe anatomical plausibility of the generated samples, we provide furtherguidance by generating vascular, airway, and lobular segmentation masks inconjunction with the CT images. The model demonstrates the capability to usetextual input and segmentation tasks to generate synthesized images. Theresults of comparative assessments indicate that our approach exhibits superiorperformance compared to the most advanced models based on GAN and diffusiontechniques, especially in accurately retaining crucial anatomical features suchas fissure lines, airways, and vascular structures. This innovation introducesnovel possibilities. This study focuses on two main objectives: (1) thedevelopment of a method for creating images based on textual prompts andanatomical components, and (2) the capability to generate new imagesconditioning on anatomical elements. The advancements in image generation canbe applied to enhance numerous downstream tasks.

MedSyn: Text-guided Anatomy-aware Synthesis of High-Fidelity 3D CT Images