• 文献检索
  • 文档翻译
  • 深度研究
  • 学术资讯
  • Suppr Zotero 插件Zotero 插件
  • 邀请有礼
  • 套餐&价格
  • 历史记录
应用&插件
Suppr Zotero 插件Zotero 插件浏览器插件Mac 客户端Windows 客户端微信小程序
定价
高级版会员购买积分包购买API积分包
服务
文献检索文档翻译深度研究API 文档MCP 服务
关于我们
关于 Suppr公司介绍联系我们用户协议隐私条款
关注我们

Suppr 超能文献

核心技术专利：CN118964589B侵权必究
粤ICP备2023148730 号-1Suppr @ 2026

文献检索

告别复杂PubMed语法，用中文像聊天一样搜索，搜遍4000万医学文献。AI智能推荐，让科研检索更轻松。

立即免费搜索

文件翻译

保留排版，准确专业，支持PDF/Word/PPT等文件格式，支持 12+语言互译。

免费翻译文档

深度研究

AI帮你快速写综述，25分钟生成高质量综述，智能提取关键信息，辅助科研写作。

立即免费体验

基于深度学习的F-氟脱氧葡萄糖正电子发射断层扫描图像质量改善：一项回顾性观察研究。

Deep learning-based image quality improvement of F-fluorodeoxyglucose positron emission tomography: a retrospective observational study.

作者信息

Tsuchiya Junichi, Yokoyama Kota, Yamagiwa Ken, Watanabe Ryosuke, Kimura Koichiro, Kishino Mitsuhiro, Chan Chung, Asma Evren, Tateishi Ukihide

机构信息

Department of Diagnostic Radiology and Nuclear Medicine, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo, 113-8510, Japan.

Canon Medical Research USA, Inc., 706 N. Deerpath Drive, Vernon Hills, IL, 60061, USA.

出版信息

EJNMMI Phys. 2021 Mar 25;8(1):31. doi: 10.1186/s40658-021-00377-4.

DOI:10.1186/s40658-021-00377-4
PMID:33765233
原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC7994470/
Abstract

BACKGROUND

Deep learning (DL)-based image quality improvement is a novel technique based on convolutional neural networks. The aim of this study was to compare the clinical value of F-fluorodeoxyglucose positron emission tomography (F-FDG PET) images obtained with the DL method with those obtained using a Gaussian filter.

METHODS

Fifty patients with a mean age of 64.4 (range, 19-88) years who underwent F-FDG PET/CT between April 2019 and May 2019 were included in the study. PET images were obtained with the DL method in addition to conventional images reconstructed with three-dimensional time of flight-ordered subset expectation maximization and filtered with a Gaussian filter as a baseline for comparison. The reconstructed images were reviewed by two nuclear medicine physicians and scored from 1 (poor) to 5 (excellent) for tumor delineation, overall image quality, and image noise. For the semi-quantitative analysis, standardized uptake values in tumors and healthy tissues were compared between images obtained using the DL method and those obtained with a Gaussian filter.

RESULTS

Images acquired using the DL method scored significantly higher for tumor delineation, overall image quality, and image noise compared to baseline (P < 0.001). The Fleiss' kappa value for overall inter-reader agreement was 0.78. The standardized uptake values in tumor obtained by DL were significantly higher than those acquired using a Gaussian filter (P < 0.001).

CONCLUSIONS

Deep learning method improves the quality of PET images.

摘要

背景

基于深度学习（DL）的图像质量改善是一种基于卷积神经网络的新技术。本研究的目的是比较使用DL方法获得的F-氟脱氧葡萄糖正电子发射断层扫描（F-FDG PET）图像与使用高斯滤波器获得的图像的临床价值。

方法

纳入2019年4月至2019年5月期间接受F-FDG PET/CT检查的50例患者，平均年龄64.4岁（范围19 - 88岁）。除了用三维飞行时间有序子集期望最大化重建并用高斯滤波器滤波作为比较基线的传统图像外，还用DL方法获得PET图像。由两名核医学医师对重建图像进行评估，在肿瘤轮廓、整体图像质量和图像噪声方面从1分（差）到5分（优）进行评分。对于半定量分析，比较使用DL方法获得的图像与使用高斯滤波器获得的图像中肿瘤和健康组织的标准化摄取值。

结果

与基线相比，使用DL方法获得的图像在肿瘤轮廓、整体图像质量和图像噪声方面得分显著更高（P < 0.001）。读者间总体一致性的Fleiss' kappa值为0.78。DL获得的肿瘤标准化摄取值显著高于使用高斯滤波器获得的值（P < 0.001）。

结论

深度学习方法可提高PET图像质量。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/49bb/7994470/6294075a9d08/40658_2021_377_Fig5_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/49bb/7994470/0cd5638dbf8e/40658_2021_377_Fig1_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/49bb/7994470/0ed68298b1ec/40658_2021_377_Fig2_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/49bb/7994470/377deec2203d/40658_2021_377_Fig3_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/49bb/7994470/b78f025688be/40658_2021_377_Fig4_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/49bb/7994470/6294075a9d08/40658_2021_377_Fig5_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/49bb/7994470/0cd5638dbf8e/40658_2021_377_Fig1_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/49bb/7994470/0ed68298b1ec/40658_2021_377_Fig2_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/49bb/7994470/377deec2203d/40658_2021_377_Fig3_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/49bb/7994470/b78f025688be/40658_2021_377_Fig4_HTML.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/49bb/7994470/6294075a9d08/40658_2021_377_Fig5_HTML.jpg

相似文献

1
Deep learning-based image quality improvement of F-fluorodeoxyglucose positron emission tomography: a retrospective observational study.基于深度学习的F-氟脱氧葡萄糖正电子发射断层扫描图像质量改善：一项回顾性观察研究。
EJNMMI Phys. 2021 Mar 25;8(1):31. doi: 10.1186/s40658-021-00377-4.
2
Enhancement of F-Fluorodeoxyglucose PET Image Quality by Deep-Learning-Based Image Reconstruction Using Advanced Intelligent Clear-IQ Engine in Semiconductor-Based PET/CT Scanners.基于深度学习的图像重建利用先进智能清晰IQ引擎在基于半导体的PET/CT扫描仪中增强F-氟脱氧葡萄糖PET图像质量
Diagnostics (Basel). 2022 Oct 15;12(10):2500. doi: 10.3390/diagnostics12102500.
3
Sub-minute acquisition with deep learning-based image filter in the diagnosis of colorectal cancers using total-body F-FDG PET/CT.基于深度学习图像滤波器的亚分钟采集在结直肠癌全身 F-FDG PET/CT 诊断中的应用
EJNMMI Res. 2023 Jul 10;13(1):66. doi: 10.1186/s13550-023-01015-z.
4
Full-count PET recovery from low-count image using a dilated convolutional neural network.使用扩张卷积神经网络从低计数图像中进行全计数PET恢复。
Med Phys. 2020 Oct;47(10):4928-4938. doi: 10.1002/mp.14402. Epub 2020 Aug 6.
5
Deep learning-assisted PET imaging achieves fast scan/low-dose examination.深度学习辅助的正电子发射断层显像（PET）成像可实现快速扫描/低剂量检查。
EJNMMI Phys. 2022 Feb 4;9(1):7. doi: 10.1186/s40658-022-00431-9.
6
The value of Bayesian penalized likelihood reconstruction for improving lesion conspicuity of malignant lung tumors on F-FDG PET/CT: comparison with ordered subset expectation maximization reconstruction incorporating time-of-flight model and point spread function correction.贝叶斯惩罚似然重建提高 F-FDG PET/CT 中恶性肺肿瘤病灶显示能力的价值：与包含飞行时间模型和点扩散函数校正的有序子集期望最大化重建的比较。
Ann Nucl Med. 2020 Apr;34(4):272-279. doi: 10.1007/s12149-020-01446-x. Epub 2020 Feb 14.
7
Image Quality and Semiquantitative Measurements on the Biograph Vision PET/CT System: Initial Experiences and Comparison with the Biograph mCT.Biograph Vision PET/CT 系统的图像质量和半定量测量：初步经验及与 Biograph mCT 的比较。
J Nucl Med. 2020 Jan;61(1):129-135. doi: 10.2967/jnumed.119.227801. Epub 2019 Jun 28.
8
Deep Learning-Based Image Quality Improvement in Digital Positron Emission Tomography for Breast Cancer.基于深度学习的乳腺癌数字正电子发射断层成像图像质量改善
Diagnostics (Basel). 2023 Feb 20;13(4):794. doi: 10.3390/diagnostics13040794.
9
More advantages in detecting bone and soft tissue metastases from prostate cancer using F-PSMA PET/CT.使用F-PSMA PET/CT检测前列腺癌骨和软组织转移方面有更多优势。
Hell J Nucl Med. 2019 Jan-Apr;22(1):6-9. doi: 10.1967/s002449910952. Epub 2019 Mar 7.
10
Standardization of image reconstruction parameters for dynamic fluorine-18-fluorodeoxyglucose positron emission tomography/computed tomography.动态氟-18-氟脱氧葡萄糖正电子发射断层扫描/计算机断层扫描图像重建参数的标准化
Nucl Med Commun. 2018 Dec;39(12):1207-1217. doi: 10.1097/MNM.0000000000000923.

引用本文的文献

1
Innovations in clinical PET image reconstruction: advances in Bayesian penalized likelihood algorithm and deep learning.临床正电子发射断层显像（PET）图像重建的创新：贝叶斯惩罚似然算法与深度学习的进展
Ann Nucl Med. 2025 Jul 18. doi: 10.1007/s12149-025-02088-7.
2
F-FDG dose reduction using deep learning-based PET reconstruction.基于深度学习的PET重建实现F-FDG剂量降低
EJNMMI Res. 2025 Jul 1;15(1):78. doi: 10.1186/s13550-025-01269-9.
3
Effects of a deep learning-based image quality enhancement method on a digital-BGO PET/CT system for F-FDG whole-body examination.

本文引用的文献

1
Convolutional neural networks for improving image quality with noisy PET data.用于利用噪声PET数据提高图像质量的卷积神经网络。
EJNMMI Res. 2020 Sep 21;10(1):105. doi: 10.1186/s13550-020-00695-1.
2
PET Image Denoising Using a Deep Neural Network Through Fine Tuning.通过微调深度学习网络实现PET图像去噪
IEEE Trans Radiat Plasma Med Sci. 2019 Mar;3(2):153-161. doi: 10.1109/TRPMS.2018.2877644. Epub 2018 Oct 23.
3
Deep Learning Based Noise Reduction for Brain MR Imaging: Tests on Phantoms and Healthy Volunteers.基于深度学习的脑磁共振成像降噪：在体模和健康志愿者上的测试。
基于深度学习的图像质量增强方法对用于F-FDG全身检查的数字化BGO PET/CT系统的影响。
EJNMMI Phys. 2025 Mar 28;12(1):29. doi: 10.1186/s40658-025-00742-7.
4
Deep learning-based time-of-flight (ToF) enhancement of non-ToF PET scans for different radiotracers.基于深度学习的非飞行时间（ToF）正电子发射断层扫描（PET）扫描对不同放射性示踪剂的飞行时间（ToF）增强
Eur J Nucl Med Mol Imaging. 2025 Feb 18. doi: 10.1007/s00259-025-07119-z.
5
Deep learning-based PET image denoising and reconstruction: a review.基于深度学习的 PET 图像去噪与重建：综述
Radiol Phys Technol. 2024 Mar;17(1):24-46. doi: 10.1007/s12194-024-00780-3. Epub 2024 Feb 6.
6
Deep convolutional neural network for differentiating between sarcoidosis and lymphoma based on [F]FDG maximum-intensity projection images.基于 [F]FDG 最大密度投影图像的深度学习卷积神经网络用于鉴别结节病与淋巴瘤。
Eur Radiol. 2024 Jan;34(1):374-383. doi: 10.1007/s00330-023-09937-x. Epub 2023 Aug 3.
7
Acquisition time reduction in pediatric Tc-DMSA planar imaging using deep learning.使用深度学习技术减少儿科 Tc-DMSA 平面显像的采集时间。
J Appl Clin Med Phys. 2023 Jun;24(6):e13978. doi: 10.1002/acm2.13978. Epub 2023 Apr 5.
8
Deep Learning-Based Image Quality Improvement in Digital Positron Emission Tomography for Breast Cancer.基于深度学习的乳腺癌数字正电子发射断层成像图像质量改善
Diagnostics (Basel). 2023 Feb 20;13(4):794. doi: 10.3390/diagnostics13040794.
9
Enhancement of F-Fluorodeoxyglucose PET Image Quality by Deep-Learning-Based Image Reconstruction Using Advanced Intelligent Clear-IQ Engine in Semiconductor-Based PET/CT Scanners.基于深度学习的图像重建利用先进智能清晰IQ引擎在基于半导体的PET/CT扫描仪中增强F-氟脱氧葡萄糖PET图像质量
Diagnostics (Basel). 2022 Oct 15;12(10):2500. doi: 10.3390/diagnostics12102500.
10
Virtual high-count PET image generation using a deep learning method.基于深度学习方法的虚拟高计数 PET 图像生成。
Med Phys. 2022 Sep;49(9):5830-5840. doi: 10.1002/mp.15867. Epub 2022 Aug 13.
Magn Reson Med Sci. 2020 Aug 3;19(3):195-206. doi: 10.2463/mrms.mp.2019-0018. Epub 2019 Sep 4.
4
Accelerated SPECT image reconstruction with FBP and an image enhancement convolutional neural network.基于傅里叶反投影（FBP）和图像增强卷积神经网络的加速单光子发射计算机断层扫描（SPECT）图像重建
EJNMMI Phys. 2019 Jul 29;6(1):14. doi: 10.1186/s40658-019-0252-0.
5
Ultra-low-dose PET reconstruction using generative adversarial network with feature matching and task-specific perceptual loss.基于特征匹配和任务特定感知损失的生成对抗网络的超低剂量 PET 重建。
Med Phys. 2019 Aug;46(8):3555-3564. doi: 10.1002/mp.13626. Epub 2019 Jun 17.
6
Deep learning reconstruction improves image quality of abdominal ultra-high-resolution CT.深度学习重建可提高腹部超高分辨率 CT 的图像质量。
Eur Radiol. 2019 Nov;29(11):6163-6171. doi: 10.1007/s00330-019-06170-3. Epub 2019 Apr 11.
7
Deep learning-based image restoration algorithm for coronary CT angiography.基于深度学习的冠状动脉 CT 血管造影图像恢复算法。
Eur Radiol. 2019 Oct;29(10):5322-5329. doi: 10.1007/s00330-019-06183-y. Epub 2019 Apr 8.
8
Ultra-Low-Dose F-Florbetaben Amyloid PET Imaging Using Deep Learning with Multi-Contrast MRI Inputs.基于多对比 MRI 输入的深度学习的超灵敏氟代脱氧葡萄糖 F-Florbetaben 淀粉样蛋白 PET 成像。
Radiology. 2019 Mar;290(3):649-656. doi: 10.1148/radiol.2018180940. Epub 2018 Dec 11.
9
Real-time cardiovascular MR with spatio-temporal artifact suppression using deep learning-proof of concept in congenital heart disease.基于深度学习的实时心血管磁共振时空伪影抑制技术：在先天性心脏病中的初步研究。
Magn Reson Med. 2019 Feb;81(2):1143-1156. doi: 10.1002/mrm.27480. Epub 2018 Sep 8.
10
Dual-energy CT in patients with colorectal cancer: Improved assessment of hypoattenuating liver metastases using noise-optimized virtual monoenergetic imaging.双能 CT 在结直肠癌患者中的应用：应用噪声优化的虚拟单能量成像技术提高低衰减肝转移瘤的评估效果。
Eur J Radiol. 2018 Sep;106:184-191. doi: 10.1016/j.ejrad.2018.07.027. Epub 2018 Jul 31.