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

Wide-angle and magnified tilted-plane holographic reconstruction using multi-segment phase encoding and spatial rotation.

Applied optics·2026
Same author

Cavernous Sinus Arteriovenous Fistula Presenting as Acute Third Nerve Palsy With Coexisting Sphenoid Sinusitis.

OTO open·2025
Same author

Elective versus involved target volume definition of stereotactic spine radiosurgery for spinal metastases: A phase II randomized clinical trial.

Neuro-oncology·2025
Same author

Erratum to: Does the Presence of Missing Data Affect the Performance of the SORG Machine-learning Algorithm for Patients With Spinal Metastasis? Development of an Internet Application Algorithm.

Clinical orthopaedics and related research·2025
Same author

Single versus multiple fraction stereotactic spine radiosurgery for spinal metastases: a prospective randomized Phase II trial.

The spine journal : official journal of the North American Spine Society·2025
Same author

Validity of the I‑FEED classification in assessing postoperative gastrointestinal impairment in patients undergoing elective lumbar spinal surgery with general anesthesia: a prospective observational study.

Perioperative medicine (London, England)·2024

相关实验视频

Updated: Jan 14, 2026

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

3.3K

Mod-SE(2):用于MRI图像中脑瘤分类和细分的几何深度学习框架.

Clara Lavita Angelina1,2,3, Fu-Ren Xiao3,4, Sunil Vyas3,5

  • 1Graduate School of Engineering Science and Technology, National Yunlin University of Science and Technology, Yunlin, 64002, Taiwan.

Journal of biomedical science
|January 13, 2026
PubMed
概括

本研究介绍了Mod-SE(2),这是一种用于脑瘤分类和细分的几何深度学习框架. 与传统的CNN相比,它显著提高了准确性和效率,有助于诊断和治疗规划.

关键词:
脑瘤分类大脑瘤的分类几何深度学习的几何深度学习这就是为什么MRI是MRI.医学成像医学成像莫德-SE (((2)) 的意思是旋转翻译的不变性

更多相关视频

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

9.4K
Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
10:25

Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping

Published on: September 25, 2019

49.3K

相关实验视频

Last Updated: Jan 14, 2026

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

3.3K
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

9.4K
Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
10:25

Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping

Published on: September 25, 2019

49.3K

科学领域:

  • 医疗成像医学成像
  • 人工智能的人工智能
  • 计算神经科学是一种神经科学.

背景情况:

  • 准确的脑瘤分类和细分对于患者的诊断和治疗至关重要.
  • 异质瘤形态对传统的卷积神经网络 (CNN) 提出了重大挑战.
  • CNNs通常缺乏旋转和转换不变性,限制了它们在各种MRI数据上的性能.

研究的目的:

  • 引入一种新的几何深度学习框架,改进的特殊欧几里德式 (Mod-SE(2)),用于增强脑瘤分析.
  • 改善空间一致性,减少瘤分类和细分中的数据增强依赖性.
  • 为了利用几何先验和保持对称性的组卷曲来进行强大的医学图像分析.

主要方法:

  • 开发了Mod-SE(2) 框架,集成几何先验和保持对称性的组卷曲.
  • 应用Mod-SE(2) 进行瘤分类 (Mod-Cls-SE(2) 和细分 (Mod-Seg-SE(2) 任务.
  • 在多个MRI和医疗图像数据集上评估了性能,与U-Net,NN U-Net,VGG16,VGG19和ResNet进行了对比.

主要成果:

  • Mod-Cls-SE(2) 实现了0.914的平均分类准确度,超过了ResNet101 (0.682) 和VGG16 (0.705).
  • 在BraTS2020上,Mod-Seg-SE(2) 获得了0.9503的子系数和0.9616的IOU,超过了U-Net (0.797) 和NN U-Net (0.815).
  • 该模型证明了推断时间的减少和强大的计算性能.

结论:

  • 通过对称感知设计,Mod-SE(2) 提高了脑瘤分析中的空间一致性,效率和可解释性.
  • 该框架在不同的瘤形状上更好地泛化,在关键指标上表现优于传统方法.
  • Mod-SE(2) 支持精确的边界划定用于神经外科规划和其他临床应用.