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

MammoDenseSegNet: A Context-Aware Deep Learning Model for Dense Tissue Segmentation in Digital Mammograms.

Journal of imaging informatics in medicine·2026
Same author

Counterfactual Reasoning for Mammogram Classification via Semantic Texture Masking.

Journal of imaging informatics in medicine·2026
Same author

Reply to: "Annual Versus Biennial Mammographic Screening" and "Breast Cancer Screening Interval: Effects of Proportions and Biases on Benefits".

Journal of clinical oncology : official journal of the American Society of Clinical Oncology·2024
Same author

Stop Training Artificial Intelligence Algorithms Now. Start Prospective Trials!

Journal of breast imaging·2024
Same author

Breast Cancer Screening Interval: Effect on Rate of Late-Stage Disease at Diagnosis and Overall Survival.

Journal of clinical oncology : official journal of the American Society of Clinical Oncology·2024
Same author

Organizational Breast Cancer Data Mart: A Solution for Assessing Outcomes of Imaging and Treatment.

JCO clinical cancer informatics·2024

相关实验视频

Updated: Jan 8, 2026

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

991

Mammo-GAN协助的深度网络培训计划用于损伤检测.

Juhun Lee1,2, Robert M Nishikawa3

  • 1Department of Radiology, The University of Pittsburgh, 200 Lothrop Street, Pittsburgh, PA, 15237, USA. leej15@upmc.edu.

Journal of imaging informatics in medicine
|December 16, 2025
PubMed
概括

使用基于Cycle-GAN的损伤模拟器 (LS) 和损伤清除器 (LR) 生成更具挑战性的边界病例显著改善了深度网络性能,用于在乳房影像和胸部X射线图像中检测损伤.

关键词:
计算机辅助检测 计算机辅助检测这是一个循环-GANAN.生成性AI是一种人工智能.伤口去除剂 伤口去除剂损伤模拟器 损伤模拟器

更多相关视频

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
07:15

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

7.3K
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

相关实验视频

Last Updated: Jan 8, 2026

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

991
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
07:15

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

7.3K
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

科学领域:

  • 医疗成像医学成像
  • 人工智能的人工智能
  • 计算机视觉 计算机视觉

背景情况:

  • 用于病变检测的深度学习模型需要广泛的训练数据,特别是边界病例,这些数据很难获得.
  • 病变检测网络的性能往往受挑战性,边界训练示例的稀缺性限制.

研究的目的:

  • 通过模拟和删除病变来增强训练数据集的新方法,以创建更具挑战性的深度学习模型.
  • 通过数据增强,通过增加边界病例数量来提高病变检测深度网络的性能.

主要方法:

  • 基于循环GAN的损伤模拟器 (LS) 和损伤清除器 (LR) 使用乳房扫描数据集开发.
  • LS在正常斑块中产生病变,LR则从异常斑块中去除病变,从而产生难以区分的病例.
  • 在不同的时代 (25%, 50%, 75%) 训练LS-LR模型以控制模拟影响,并使用增强数据重新训练ResNet18模型.

主要成果:

  • 在50%/75%的训练时段用LS-LR增强的数据重新训练ResNet18显著改善了乳房扫描测试组的病变检测性能 (AUC=0.901) 与基线 (AUC=0.870) 相比.
  • 外部验证证明了可通用性,AUC在独立的乳腺造影数据集 (0.866比0.839) 和胸部X射线数据集 (0.975比0.964) 上有所改善.

结论:

  • 拟议的LS-LR模型有效地将现有的医学成像数据转化为边界病例,从而提高基于深度学习的病变检测的稳定性和性能.
  • 这种数据增强策略提供了一种有前途的方法,可以提高乳房造影和胸部X射线解释的诊断准确性,解决边界病例可用性有限的挑战.