• 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

深度学习用于未校正扫描透射电子显微镜中的亚埃分辨率成像。

Deep learning for sub-ångström-resolution imaging in uncorrected scanning transmission electron microscopy.

作者信息

Qiu Zanlin, Meng Yuan, Li Junxian, Hong Yanhui, Li Ning, Han Xiaocang, Liang Yu, Cheng Wing Ni, Ke Guolin, Zhang Linfeng, E Weinan, Zhao Xiaoxu, Zhang Jin

机构信息

School of Materials Science and Engineering, Peking University, Beijing 100871, China.

DP Technology, Beijing 100080, China.

出版信息

Natl Sci Rev. 2025 Jun 5;12(8):nwaf235. doi: 10.1093/nsr/nwaf235. eCollection 2025 Aug.

DOI:10.1093/nsr/nwaf235
PMID:40755523
原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC12315106/
Abstract

Achieving sub-ångström resolution has long been restricted to sophisticated aberration-corrected scanning transmission electron microscopy (AC-STEM). Recent advances in computational super-resolution techniques, such as deconvolution and electron ptychography, have enabled uncorrected STEM to achieve sub-ångström resolution without the need for delicate aberration correctors. However, these methods have strict requirements for sample thickness and thus have yet to be widely implemented. In this study, we introduce SARDiffuse-a deep-learning diffusion model designed to enhance spatial resolution and correct the noise level of uncorrected STEM images. Trained with experimental AC-STEM data, SARDiffuse has the capability to restore high-frequency information of STEM images, enabling sub-ångström resolution in an uncorrected microscope. We demonstrate the effectiveness of the model on representative materials, including silicon, strontium titanate and gallium nitride, achieving substantial improvements (<1 Å) in spatial resolution. Detailed statistical analysis confirms that SARDiffuse reliably preserves atomic positions, demonstrating that it is a powerful tool for high-precision material characterization. Furthermore, SARDiffuse effectively mitigates spherical-aberration-induced artifacts, outperforming current methods in artifact correction. Meanwhile, the background information of images, such as thickness variation or carbon contamination distribution, is also preserved. This work highlights the potential of deep learning to realize sub-ångström-resolution imaging in the uncorrected electron microscope, offering a cost-effective alternative to delicate AC-STEM when imaging conventional single crystals.

摘要

长期以来,实现亚埃分辨率一直局限于复杂的像差校正扫描透射电子显微镜(AC-STEM)。反卷积和电子叠层成像等计算超分辨率技术的最新进展,使未校正的STEM能够在无需精密像差校正器的情况下实现亚埃分辨率。然而,这些方法对样品厚度有严格要求,因此尚未得到广泛应用。在本研究中,我们引入了SARDiffuse——一种深度学习扩散模型,旨在提高空间分辨率并校正未校正STEM图像的噪声水平。通过实验AC-STEM数据进行训练,SARDiffuse有能力恢复STEM图像的高频信息,从而在未校正的显微镜中实现亚埃分辨率。我们在包括硅、钛酸锶和氮化镓在内的代表性材料上证明了该模型的有效性,在空间分辨率上实现了大幅提升(<1 Å)。详细的统计分析证实,SARDiffuse能够可靠地保留原子位置,表明它是高精度材料表征的有力工具。此外,SARDiffuse有效地减轻了球差引起的伪像,在伪像校正方面优于当前方法。同时,图像的背景信息,如厚度变化或碳污染分布,也得以保留。这项工作突出了深度学习在未校正电子显微镜中实现亚埃分辨率成像的潜力,为常规单晶成像时提供了一种经济高效的替代精密AC-STEM的方法。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/14fc/12315106/2855038e4763/nwaf235fig4.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/14fc/12315106/bbf34771dd25/nwaf235fig1.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/14fc/12315106/f69761d16f6d/nwaf235fig2.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/14fc/12315106/1b869b408980/nwaf235fig3.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/14fc/12315106/2855038e4763/nwaf235fig4.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/14fc/12315106/bbf34771dd25/nwaf235fig1.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/14fc/12315106/f69761d16f6d/nwaf235fig2.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/14fc/12315106/1b869b408980/nwaf235fig3.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/14fc/12315106/2855038e4763/nwaf235fig4.jpg

相似文献

1
Deep learning for sub-ångström-resolution imaging in uncorrected scanning transmission electron microscopy.深度学习用于未校正扫描透射电子显微镜中的亚埃分辨率成像。
Natl Sci Rev. 2025 Jun 5;12(8):nwaf235. doi: 10.1093/nsr/nwaf235. eCollection 2025 Aug.
2
Prescription of Controlled Substances: Benefits and Risks管制药品的处方:益处与风险
3
Short-Term Memory Impairment短期记忆障碍
4
Management of urinary stones by experts in stone disease (ESD 2025).结石病专家对尿路结石的管理(2025年结石病专家共识)
Arch Ital Urol Androl. 2025 Jun 30;97(2):14085. doi: 10.4081/aiua.2025.14085.
5
The Black Book of Psychotropic Dosing and Monitoring.《精神药物剂量与监测黑皮书》
Psychopharmacol Bull. 2024 Jul 8;54(3):8-59.
6
Cost-effectiveness of using prognostic information to select women with breast cancer for adjuvant systemic therapy.利用预后信息为乳腺癌患者选择辅助性全身治疗的成本效益
Health Technol Assess. 2006 Sep;10(34):iii-iv, ix-xi, 1-204. doi: 10.3310/hta10340.
7
Preserving noise texture through training data curation for deep learning denoising of high-resolution cardiac EID-CT.通过训练数据精选来保留噪声纹理,用于高分辨率心脏EID-CT的深度学习去噪
Med Phys. 2025 Jul;52(7):e17938. doi: 10.1002/mp.17938.
8
Distilling knowledge from graph neural networks trained on cell graphs to non-neural student models.从在细胞图上训练的图神经网络中提取知识,用于非神经学生模型。
Sci Rep. 2025 Aug 10;15(1):29274. doi: 10.1038/s41598-025-13697-7.
9
Electrophoresis电泳
10
The clinical effectiveness and cost-effectiveness of enzyme replacement therapy for Gaucher's disease: a systematic review.戈谢病酶替代疗法的临床疗效和成本效益:一项系统评价。
Health Technol Assess. 2006 Jul;10(24):iii-iv, ix-136. doi: 10.3310/hta10240.

本文引用的文献

1
Emittance minimization for aberration correction I: Aberration correction of an electron microscope without knowing the aberration coefficients.用于像差校正的发射度最小化I:在不知道像差系数的情况下对电子显微镜进行像差校正。
Ultramicroscopy. 2025 Jul;273:114137. doi: 10.1016/j.ultramic.2025.114137. Epub 2025 Apr 5.
2
Visualization of oxygen vacancies and self-doped ligand holes in LaNiO.LaNiO 中氧空位和自掺杂配体空穴的可视化
Nature. 2024 Jun;630(8018):847-852. doi: 10.1038/s41586-024-07482-1. Epub 2024 Jun 5.
3
Diffusion-based deep learning method for augmenting ultrastructural imaging and volume electron microscopy.
基于扩散的深度学习方法用于增强超微结构成像和体积电子显微镜技术。
Nat Commun. 2024 Jun 1;15(1):4677. doi: 10.1038/s41467-024-49125-z.
4
Diffusion models in bioinformatics and computational biology.生物信息学和计算生物学中的扩散模型。
Nat Rev Bioeng. 2024 Feb;2(2):136-154. doi: 10.1038/s44222-023-00114-9. Epub 2023 Oct 27.
5
Microstructure reconstruction of 2D/3D random materials via diffusion-based deep generative models.基于扩散的深度生成模型对二维/三维随机材料的微观结构重建
Sci Rep. 2024 Feb 29;14(1):5041. doi: 10.1038/s41598-024-54861-9.
6
Achieving sub-0.5-angstrom-resolution ptychography in an uncorrected electron microscope.在未校正的电子显微镜中实现亚0.5埃分辨率的叠层成像技术。
Science. 2024 Feb 23;383(6685):865-870. doi: 10.1126/science.adl2029. Epub 2024 Feb 22.
7
Local-orbital ptychography for ultrahigh-resolution imaging.用于超高分辨率成像的局域轨道叠层成像技术
Nat Nanotechnol. 2024 May;19(5):612-617. doi: 10.1038/s41565-023-01595-w. Epub 2024 Jan 29.
8
Learning motifs and their hierarchies in atomic resolution microscopy.在原子分辨率显微镜中学习基序及其层次结构。
Sci Adv. 2022 Apr 15;8(15):eabk1005. doi: 10.1126/sciadv.abk1005. Epub 2022 Apr 13.
9
Reprint of: Automated geometric aberration correction for large-angle illumination STEM.重印:大角度照明扫描透射电子显微镜的自动几何像差校正
Ultramicroscopy. 2021 Dec;231:113410. doi: 10.1016/j.ultramic.2021.113410. Epub 2021 Oct 28.
10
High-endurance micro-engineered LaB nanowire electron source for high-resolution electron microscopy.用于高分辨率电子显微镜的高性能微纳加工镧硼纳米线电子源。
Nat Nanotechnol. 2022 Jan;17(1):21-26. doi: 10.1038/s41565-021-00999-w. Epub 2021 Nov 8.