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

Automated segmentation of pituitary adenomas, pituitary gland, and internal carotid arteries on routine coronal contrast-enhanced T1-weighted MRI: a single-sequence feasibility study.

Frontiers in endocrinology·2026
Same author

Customized three dimensionally printed palatal stents: A randomized controlled clinical trial.

Clinical advances in periodontics·2026
Same author

Methodology for Evaluating Behavior of Reinforced Concrete Slabs in Temporary Traffic Bridge Systems over Uncured Cement Concrete Pavements Using Small-Scale Experimental Slabs.

Materials (Basel, Switzerland)·2026
Same author

The C2H2-type zinc finger protein SlHair5 acts as a regulatory node coordinating trichome development and leaf morphogenesis in tomato.

Plant cell reports·2026
Same author

LC2 screws may significantly increase fixation stability when compared with plate osteosynthesis in type IIIa fragility fractures of the pelvis: A biomechanical comparison study.

Injury·2026
Same author

Mutation of STAY-GREEN 1 in tomato increases volatile organic compounds during fruit ripening.

Plant & cell physiology·2026

相关实验视频

Updated: Jul 18, 2025

Author Spotlight: Three-Dimensional Cephalometric Landmark Annotation Demonstration on Human Cone Beam Computed Tomography Scans
10:23

Author Spotlight: Three-Dimensional Cephalometric Landmark Annotation Demonstration on Human Cone Beam Computed Tomography Scans

Published on: September 8, 2023

2.8K

使用深度强化学习的自动头脑测量地标检测

Woojae Hong1, Seong-Min Kim1, Joongyeon Choi1

  • 1Department of Biomechatronic Engineering, Sungkyunkwan University, Suwon, Gyeonggi.

The Journal of craniofacial surgery
|August 25, 2023
PubMed
概括

本研究介绍了深度Q网络 (DQN) 和双深度Q网络 (DDQN) 用于自动化脑测量地标检测. 这些强化学习方法达到临床上接受的准确性,显示出在牙科分析中实际使用的潜力.

更多相关视频

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization
05:49

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization

Published on: February 23, 2024

884
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

570

相关实验视频

Last Updated: Jul 18, 2025

Author Spotlight: Three-Dimensional Cephalometric Landmark Annotation Demonstration on Human Cone Beam Computed Tomography Scans
10:23

Author Spotlight: Three-Dimensional Cephalometric Landmark Annotation Demonstration on Human Cone Beam Computed Tomography Scans

Published on: September 8, 2023

2.8K
Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization
05:49

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization

Published on: February 23, 2024

884
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

570

科学领域:

  • 医疗成像医学成像
  • 人工智能的人工智能
  • 生物医学工程 生物医学工程

背景情况:

  • 精确的脑力测量地标检测对于牙科分析,诊断和手术规划至关重要.
  • 现有的自动化方法有局限性,强化学习还没有应用于此任务.
  • 这项研究开创了深度Q网络 (DQN) 和双深度Q网络 (DDQN) 为此目的的应用.

研究的目的:

  • 应用和评估深度Q网络 (DQN) 和双深度Q网络 (DDQN) 用于自动化脑测地标检测.
  • 用现有方法比较基于DQN的网络的性能.
  • 评估这些新方法的临床适用性.

主要方法:

  • 实施深度Q网络 (DQN) 和双深度Q网络 (DDQN) 算法用于地标检测.
  • 使用IEEE生物医学成像国际研讨会 (ISBI) 2015挑战数据集进行评估.
  • 在500名患者的临床数据集上进行验证.

主要成果:

  • 基于DQN的网络实现了19个地标的平均半径误差低于2毫米,符合临床标准.
  • 没有数据增强或预处理,DQN方法表现出高准确度.
  • 在500名患者的临床数据集上,DQN和DDQN方法在2毫米范围内分别实现了67.33%和66.04%的成功检测率.

结论:

  • 深度Q网络 (DQN) 和双深度Q网络 (DDQN) 是有效的自动头脑测量地标检测.
  • 这些强化学习方法证明了牙科临床应用的可行性和潜力.
  • 这些方法在没有大量数据预处理的情况下实现了临床接受的准确性.