Deep learning in electron microscopy(EM)data analysis is predominantly supervised,relying on manually labeled data.This dependence limits scalability and slows the development of highthroughput EM characterization of ...Deep learning in electron microscopy(EM)data analysis is predominantly supervised,relying on manually labeled data.This dependence limits scalability and slows the development of highthroughput EM characterization of materials.While simulation-based approaches provide an alternative,they often struggle with morphological heterogeneity,contrast complexity,and experimental artifacts,reducing their real-world effectiveness.Weintroduce EMcopilot,a closed-loop generative learning framework that enables label-free EM segmentation.EMcopilot leverages the general vision model to extract morphological priors and employs a conditional generative adversarial network to generate contrast-aware images.An EM-specificdomain adapter further enhances realism by modeling key microscope-specific perturbations.Benchmark results show that EMcopilot-trained models not only achieve segmentation accuracy comparable to human-annotated models but also outperform them in detecting nanoparticles in poor-contrast regions and spatially clustered configurations,overcoming inherent human biases in annotation.By illustrating how generative models distill and transform complex EM features into a robust training resource in a self-supervised manner,EMcopilot provides a scalable solution for automated microscopy analysis.展开更多
Correction:Cell Regeneration 14,8(2025)https://doi.org/10.1186/s13619-025-00227-z Following publication of the original article(Hu et al.2025),the authors reported an error in Fig.1E,the FACS data of surface markers C...Correction:Cell Regeneration 14,8(2025)https://doi.org/10.1186/s13619-025-00227-z Following publication of the original article(Hu et al.2025),the authors reported an error in Fig.1E,the FACS data of surface markers CD11b and CD18 on iNEUs were mistakenly duplicated.Upon checking the original raw data,this error was caused by accidentally duplicating the same picture when formatting the figure.展开更多
基金support from the National Research Foundation(NRF)Singapore,under its NRF Fellowship(NRF-NRFF11-2019-0002)Singapore Low-Carbon Energy Research Program Funding Initiative hosted under A*STAR(Grant No.U2305d4003).We thank Dr.Jen-It Wong,Senior Application Engineer of JEOL Asia Pte Ltd,for his help in configuring computer connections for the in-situ experiments.
文摘Deep learning in electron microscopy(EM)data analysis is predominantly supervised,relying on manually labeled data.This dependence limits scalability and slows the development of highthroughput EM characterization of materials.While simulation-based approaches provide an alternative,they often struggle with morphological heterogeneity,contrast complexity,and experimental artifacts,reducing their real-world effectiveness.Weintroduce EMcopilot,a closed-loop generative learning framework that enables label-free EM segmentation.EMcopilot leverages the general vision model to extract morphological priors and employs a conditional generative adversarial network to generate contrast-aware images.An EM-specificdomain adapter further enhances realism by modeling key microscope-specific perturbations.Benchmark results show that EMcopilot-trained models not only achieve segmentation accuracy comparable to human-annotated models but also outperform them in detecting nanoparticles in poor-contrast regions and spatially clustered configurations,overcoming inherent human biases in annotation.By illustrating how generative models distill and transform complex EM features into a robust training resource in a self-supervised manner,EMcopilot provides a scalable solution for automated microscopy analysis.
文摘Correction:Cell Regeneration 14,8(2025)https://doi.org/10.1186/s13619-025-00227-z Following publication of the original article(Hu et al.2025),the authors reported an error in Fig.1E,the FACS data of surface markers CD11b and CD18 on iNEUs were mistakenly duplicated.Upon checking the original raw data,this error was caused by accidentally duplicating the same picture when formatting the figure.
