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

Non-equilibrium in the Cell01:16

Non-equilibrium in the Cell

An important concept in studying metabolism and energy is that of chemical equilibrium. Most chemical reactions are reversible. They can proceed in both directions, releasing energy into their environment in one direction, and absorbing it from the environment in the other direction. The same is true for the chemical reactions involved in cell metabolism, such as the breaking down and building up of proteins into and from individual amino acids, respectively. Reactants within a closed system...

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

Additional Dose Reduction Potential of Vendor-Agnostic Deep Learning Model: A Phantom Study.

Journal of the Korean Society of Radiology·2026
Same author

Computed Tomography-Based Artificial Intelligence Analysis of Myosteatosis and Its Effect on Survival in Fournier Gangrene.

The world journal of men's health·2026
Same author

Deep Learning-Based Augmented Contrast-Enhancement and Denoising for Reduced-Iodine and Low-Radiation 70-kVp Cerebral CT Angiography: A Prospective Study.

Korean journal of radiology·2026
Same author

Case Report: Pyogranulomatous Panniculitis Mimicking an Invasive Subcutaneous Mass with Muscular Pseudo-Invasion in a Dog.

Veterinary sciences·2026
Same author

Versatile Reactive Gradient Block Copolymer System for Highly Robust Nanopatterns With Spontaneous Vertical Orientation.

Small (Weinheim an der Bergstrasse, Germany)·2025
Same author

Unraveling the relative impact of material and optical stochastic effects on EUV LWR.

Scientific reports·2025

相关实验视频

Updated: Jun 25, 2026

Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
06:48

Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images

Published on: January 7, 2019

8.8K

自动审计和自我纠正算法用于使用MeshCNN基于按需生成AI的分段幻觉.

Sihwan Kim1,2, Changmin Park1,2, Gwanghyeon Jeon2

  • 1Department of Applied Bioengineering, Graduate School of Convergence Science and Technology, Seoul National University, Seoul 08826, Republic of Korea.

Bioengineering (Basel, Switzerland)
|January 24, 2025
PubMed
概括

我们开发了一个自动化算法来检测和修复医疗图像中的Seg-Hallucinations,提高AI细分精度,而不需要地面真相数据.

关键词:
人工智能审计AI审计审计隔离幻觉 隔离幻觉异常查 异常查 异常查细分化 细分化的细分化不确定性是一种不确定性.

更多相关视频

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

2.6K
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
04:25

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

2.2K

相关实验视频

Last Updated: Jun 25, 2026

Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
06:48

Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images

Published on: January 7, 2019

8.8K
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

2.6K
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
04:25

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

2.2K

科学领域:

  • 医疗成像医学成像
  • 人工智能的人工智能
  • 深度学习 (Deep Learning) 是一种深度学习.

背景情况:

  • 深度学习模型在医学图像细分方面表现出色,但在概括方面扎,并可能产生不现实的Seg-Hallucinations.
  • 细分幻觉导致不准确的定量分析和失去关键的成像生物标记信息.
  • 目前用于审计或纠正Seg-幻觉的现有方法是有限的.

研究的目的:

  • 提出一个自动化的Seg-幻觉监测和校正 (ASHSC) 算法.
  • 为了解决Seg-幻觉,只使用CT图像中的3D器官面具信息,没有基本真相.
  • 提高基于深度学习的医疗图像细分的可靠性和效率.

主要方法:

  • 开发了一个双阶段的,按需的ASHSC算法,使用基于网状卷积神经网络和生成AI.
  • 利用CT扫描中的3D器官面具信息进行公开可用的数据集的培训和评估.
  • 基于质量级别 (SQ级别) 的员工细分监测和按需纠正策略.

主要成果:

  • 监测阶段实现了高性能,AUROC为0.94 ± 0.01,灵敏度为0.82 ± 0.03,特异性为0.90 ± 0.01,PPV为0.92 ± 0.01.
  • 与单独的AI细分相比,按需校正阶段显著改善了所有相似度指标 (子得分,体积误差,表面距离,豪斯多夫距离).
  • 该ASHSC算法证明了有效的Seg-幻觉处理没有地面真相,为不确定性区域提供3D指导.

结论:

  • 在基于深度学习的医学图像细分中,ASHSC算法有效地审计和纠正Seg-Hallucinations.
  • 这种方法消除了对地面真相数据的需求,提高了实用性和效率.
  • ASHSC算法推进了自动审计和校正方法,提高了医学成像分析的可靠性.