Search research articles

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 Journal Methods Collections JoVE Encyclopedia of Experiments 存档

教育

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual 教师资源中心教师网站

使用条款与条件

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

Genome-Wide Identification and Expression of the Mulberry PLA Family Under Drought and Salinity.

Biology·2026

Same author

Long-read sequencing identifies complex structural variants in DMD patients.

BMC medical genomics·2026

Same author

Flow-QSM: bridging learned priors and physical models for quantitative susceptibility mapping.

Physics in medicine and biology·2026

Same author

FeaSion decodes the regulatory landscape and functional diversity of RNA polymerase II CTD phosphorylation.

Science advances·2025

Same author

Rapid whole-brain T<sub>2</sub>* and susceptibility mapping using 3D multiple overlapping-echo detachment acquisition and missing modality synthesis embedded simulation.

Magnetic resonance in medicine·2025

Same author

Identification of an F8 complex recombination in Chinese hemophilia a patient using long-read sequencing and optical genome mapping.

BMC medical genomics·2025

Search research articles

相关实验视频

Updated: Mar 7, 2026

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

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

Published on: November 30, 2022

自主监督的基于预培训的混合网络,用于深灰质核细分深灰质核.

Yang Deng¹, Jiaxiu Xi², Zhong Chen^2,1

¹Institute of Artificial Intelligence, Xiamen University, Xiamen, 361000, China.

Magnetic resonance letters

|March 6, 2026

概括

这项研究引入了一种新的混合网络 (CTNet),用于在脑MRI中对深灰质核进行细分. 自主监督的预培训显著提高了细分精度,特别是在有限的标记数据下.

关键词:

深灰色物质核的细分深灰色物质核的细分蒙面特征重建的重建旋转预测旋转预测自主监督学习学习变压器变压器变压器

更多相关视频

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Published on: November 11, 2022

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

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

Published on: January 7, 2019

相关实验视频

Last Updated: Mar 7, 2026

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

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

Published on: November 30, 2022

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Deep Learning-Based Segmentation of Cryo-Electron Tomograms

Published on: November 11, 2022

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

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

Published on: January 7, 2019

科学领域:

医学图像分析医学图像分析
人工智能在医学中的应用
神经成像是一种神经成像.

背景情况:

精确细分深灰质核对于诊断和治疗神经系统疾病至关重要.
监督学习方法需要广泛的标记数据集,这些数据集对于医学成像来说很难和耗时.
卷积神经网络 (CNN) 在捕获对复杂的细分任务至关重要的远程依赖性方面存在局限性.

研究的目的:

开发一种使用新型混合网络进行深灰质核细分的改进方法.
利用自我监督学习来减少对大型标记数据集的依赖.
通过结合本地和全球特征提取来提高细分性能.

主要方法:

提出了一个CNN-变压器混合网络 (CTNet),将3D CNN集成为本地特征和视觉变压器 (ViT) 集成为全球环境.
实施了自主监督学习 (SSL) 策略,涉及旋转预测和掩盖特征重建,用于在未标记的数据上预训练CTNet.
在3T和7T人类大脑MRI数据集上评估了该方法.

主要成果:

与现有模型相比,拟议的CTNet表现出优越的细分性能.
自主监督的预训练策略优于其他先进的SSL方法.
即使只有一个标记样本,预训练的CTNet在随机初始化的网络中,在子相似系数 (DSC) 中也显示了8.4%的改善.

结论:

开发的CTNet采用了新的SSL方法进行预训练,为深灰质核细分提供了强大而高效的解决方案.
这种方法显著提高了细分的准确性,特别是在低数据的制度,解决医疗图像分析的一个关键挑战.
混合架构有效地捕获本地和全球信息,从而在神经成像细分方面实现了最先进的性能.