Jove
Visualize
联系我们
JoVE
x logofacebook logolinkedin logoyoutube logo
关于 JoVE
概览领导团队博客JoVE 帮助中心
作者
出版流程编辑委员会范围与政策同行评审常见问题投稿
图书馆员
用户评价订阅访问资源图书馆顾问委员会常见问题
研究
JoVE JournalMethods CollectionsJoVE Encyclopedia of Experiments存档
教育
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab Manual教师资源中心教师网站
使用条款与条件
隐私政策
政策

相关概念视频

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序
Same author

Multisource Machine Learning Model for Detecting Referral-Warranted Retinopathy of Prematurity.

Ophthalmology science·2026
Same author

Essence: A benchmarking-validated transformer framework for early diagnosis of Parkinson's disease using cerebrospinal fluid protein biomarkers.

International journal of biological macromolecules·2026
Same author

In silico discovery of umami peptides: From sequence-based intelligent models to structure-function mechanistic insight.

Current research in food science·2026
Same author

Aegis: a transformer-based deep learning framework for the accurate identification of anticancer peptides.

BMC biology·2026
Same author

PlantAMP: A fine-tuned protein large language model for plant antimicrobial peptide prediction.

Plant communications·2025
Same author

From multi-omics to deep learning: advances in cfDNA-based liquid biopsy for multi-cancer screening.

Biomarker research·2025

相关实验视频

Updated: Sep 18, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

3.0K

基于引导图像过的低剂量CT图像消除方法的自主监督学习.

Yu He1, Xinwei Luo1, Chengxiang Wang1

  • 1School of Mathematical Sciences, Chongqing Normal University, ChongQing 401331, People's Republic of China.

Physics in medicine and biology
|June 25, 2025
PubMed
概括
此摘要是机器生成的。

这项研究引入了一种自我监督的深度学习方法,用于在不需要配对数据的情况下消除低剂量计算机断层扫描 (LDCT) 图像. 该方法使用引导图像过来生成伪标签,显著提高图像质量并保留结构细节.

关键词:
注意力门的注意力门.有导向的图像过.图像去色化 图像去色化低剂量CTCT可以使用.自主监督学习学习

更多相关视频

X-ray Dose Reduction through Adaptive Exposure in Fluoroscopic Imaging
08:30

X-ray Dose Reduction through Adaptive Exposure in Fluoroscopic Imaging

Published on: September 11, 2011

14.5K
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
14:08

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

42.8K

相关实验视频

Last Updated: Sep 18, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

3.0K
X-ray Dose Reduction through Adaptive Exposure in Fluoroscopic Imaging
08:30

X-ray Dose Reduction through Adaptive Exposure in Fluoroscopic Imaging

Published on: September 11, 2011

14.5K
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
14:08

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

42.8K

科学领域:

  • 医疗成像医学成像
  • 深度学习 (Deep Learning) 是一种深度学习.
  • 图像处理 图像处理

背景情况:

  • 低剂量计算机断层扫描 (LDCT) 成像对于减少辐射暴露至关重要,但导致显著的图像噪声.
  • 现有的深度学习解密方法通常需要配对的正常剂量和低剂量数据,这些数据很难获得.
  • 需要有效的LDCT拒绝技术,而不依赖于配对训练数据集.

研究的目的:

  • 为LDCT图像开发一种新的自我监督的染方法,消除对正常剂量数据配对的需求.
  • 通过在剩余网络架构中集成注意力门 (AG) 机制来增强无声化性能.

主要方法:

  • 一个自主监督的Denoising框架,使用引导图像过 (GIF) 来从LDCT图像中生成伪标签.
  • 在残余网络的解码器中实施注意力门 (AG) 机制,以改善功能聚焦和降噪.
  • 仅使用LDCT图像来训练网络,以学习输入和生成伪标签之间的噪声分布.

主要成果:

  • 拟议的方法表现出高于最先进的无监督,以变压器为基础和后处理方法相比,更高的脱光性能.
  • 视觉质量和定量指标都显示了无色的LDCT图像的显著改善.
  • 废弃研究证实了嵌入式AG机制的网络架构的最佳性能.

结论:

  • 自主监督学习与GIF相结合,有效地使LDCT能够在没有配对数据的情况下进行denoising.
  • 集成的注意力门机制显著提高了消噪能力,改善了特征聚焦和结构保存.
  • 这种方法为临床环境中高质量的LDCT图像重建提供了可行的解决方案.