Pub Date : 2025-03-03DOI: 10.1038/s41592-025-02611-8
Erik Serrano, Srinivas Niranj Chandrasekaran, Dave Bunten, Kenneth I Brewer, Jenna Tomkinson, Roshan Kern, Michael Bornholdt, Stephen J Fleming, Ruifan Pei, John Arevalo, Hillary Tsang, Vincent Rubinetti, Callum Tromans-Coia, Tim Becker, Erin Weisbart, Charlotte Bunne, Alexandr A Kalinin, Rebecca Senft, Stephen J Taylor, Nasim Jamali, Adeniyi Adeboye, Hamdah Shafqat Abbasi, Allen Goodman, Juan C Caicedo, Anne E Carpenter, Beth A Cimini, Shantanu Singh, Gregory P Way
Advances in high-throughput microscopy have enabled the rapid acquisition of large numbers of high-content microscopy images. Next, whether by deep learning or classical algorithms, image analysis pipelines commonly produce single-cell features. To process these single cells for downstream applications, we present Pycytominer, a user-friendly, open-source Python package that implements the bioinformatics steps key to image-based profiling. We demonstrate Pycytominer's usefulness in a machine-learning project to predict nuisance compounds that cause undesirable cell injuries.
{"title":"Reproducible image-based profiling with Pycytominer.","authors":"Erik Serrano, Srinivas Niranj Chandrasekaran, Dave Bunten, Kenneth I Brewer, Jenna Tomkinson, Roshan Kern, Michael Bornholdt, Stephen J Fleming, Ruifan Pei, John Arevalo, Hillary Tsang, Vincent Rubinetti, Callum Tromans-Coia, Tim Becker, Erin Weisbart, Charlotte Bunne, Alexandr A Kalinin, Rebecca Senft, Stephen J Taylor, Nasim Jamali, Adeniyi Adeboye, Hamdah Shafqat Abbasi, Allen Goodman, Juan C Caicedo, Anne E Carpenter, Beth A Cimini, Shantanu Singh, Gregory P Way","doi":"10.1038/s41592-025-02611-8","DOIUrl":"10.1038/s41592-025-02611-8","url":null,"abstract":"<p><p>Advances in high-throughput microscopy have enabled the rapid acquisition of large numbers of high-content microscopy images. Next, whether by deep learning or classical algorithms, image analysis pipelines commonly produce single-cell features. To process these single cells for downstream applications, we present Pycytominer, a user-friendly, open-source Python package that implements the bioinformatics steps key to image-based profiling. We demonstrate Pycytominer's usefulness in a machine-learning project to predict nuisance compounds that cause undesirable cell injuries.</p>","PeriodicalId":18981,"journal":{"name":"Nature Methods","volume":" ","pages":""},"PeriodicalIF":36.1,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143542546","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-27DOI: 10.1038/s41592-025-02612-7
Austin E Y T Lefebvre, Gabriel Sturm, Ting-Yu Lin, Emily Stoops, Magdalena Preciado López, Benjamin Kaufmann-Malaga, Kayley Hake
Cellular organelles undergo constant morphological changes and dynamic interactions that are fundamental to cell homeostasis, stress responses and disease progression. Despite their importance, quantifying organelle morphology and motility remains challenging due to their complex architectures, rapid movements and the technical limitations of existing analysis tools. Here we introduce Nellie, an automated and unbiased pipeline for segmentation, tracking and feature extraction of diverse intracellular structures. Nellie adapts to image metadata and employs hierarchical segmentation to resolve sub-organellar regions, while its radius-adaptive pattern matching enables precise motion tracking. Through a user-friendly Napari-based interface, Nellie enables comprehensive organelle analysis without coding expertise. We demonstrate Nellie's versatility by unmixing multiple organelles from single-channel data, quantifying mitochondrial responses to ionomycin via graph autoencoders and characterizing endoplasmic reticulum networks across cell types and time points. This tool addresses a critical need in cell biology by providing accessible, automated analysis of organelle dynamics.
{"title":"Nellie: automated organelle segmentation, tracking and hierarchical feature extraction in 2D/3D live-cell microscopy.","authors":"Austin E Y T Lefebvre, Gabriel Sturm, Ting-Yu Lin, Emily Stoops, Magdalena Preciado López, Benjamin Kaufmann-Malaga, Kayley Hake","doi":"10.1038/s41592-025-02612-7","DOIUrl":"10.1038/s41592-025-02612-7","url":null,"abstract":"<p><p>Cellular organelles undergo constant morphological changes and dynamic interactions that are fundamental to cell homeostasis, stress responses and disease progression. Despite their importance, quantifying organelle morphology and motility remains challenging due to their complex architectures, rapid movements and the technical limitations of existing analysis tools. Here we introduce Nellie, an automated and unbiased pipeline for segmentation, tracking and feature extraction of diverse intracellular structures. Nellie adapts to image metadata and employs hierarchical segmentation to resolve sub-organellar regions, while its radius-adaptive pattern matching enables precise motion tracking. Through a user-friendly Napari-based interface, Nellie enables comprehensive organelle analysis without coding expertise. We demonstrate Nellie's versatility by unmixing multiple organelles from single-channel data, quantifying mitochondrial responses to ionomycin via graph autoencoders and characterizing endoplasmic reticulum networks across cell types and time points. This tool addresses a critical need in cell biology by providing accessible, automated analysis of organelle dynamics.</p>","PeriodicalId":18981,"journal":{"name":"Nature Methods","volume":" ","pages":""},"PeriodicalIF":36.1,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143523616","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-27DOI: 10.1038/s41592-025-02613-6
{"title":"Universal organelle analyzer enables complex analyses with just one click.","authors":"","doi":"10.1038/s41592-025-02613-6","DOIUrl":"https://doi.org/10.1038/s41592-025-02613-6","url":null,"abstract":"","PeriodicalId":18981,"journal":{"name":"Nature Methods","volume":" ","pages":""},"PeriodicalIF":36.1,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143523904","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The phosphatase and tensin homolog (PTEN) is a vital protein that maintains an inhibitory brake for cellular proliferation and growth. Accordingly, PTEN loss-of-function mutations are associated with a broad spectrum of human pathologies. Despite its importance, there is currently no method to directly monitor PTEN activity with cellular specificity within intact biological systems. Here we describe the development of a FRET-based biosensor using PTEN conformation as a proxy for the PTEN activity state, for two-photon fluorescence lifetime imaging microscopy. We identify a point mutation that allows the monitoring of PTEN activity with minimal interference to endogenous PTEN signaling. We demonstrate imaging of PTEN activity in cell lines, intact Caenorhabditis elegans and in the mouse brain. Finally, we develop a red-shifted sensor variant that allows us to identify cell-type-specific PTEN activity in excitatory and inhibitory cortical cells. In summary, our approach enables dynamic imaging of PTEN activity in vivo with unprecedented spatial and temporal resolution.
{"title":"Genetically encoded biosensor for fluorescence lifetime imaging of PTEN dynamics in the intact brain.","authors":"Tomer Kagan, Matan Gabay, Aasha Meenakshisundaram, Yossi Levi, Sharbel Eid, Nikol Malchenko, Maya Maman, Anat Nitzan, Luca Ravotto, Ronen Zaidel-Bar, Britta Johanna Eickholt, Maayan Gal, Tal Laviv","doi":"10.1038/s41592-025-02610-9","DOIUrl":"https://doi.org/10.1038/s41592-025-02610-9","url":null,"abstract":"<p><p>The phosphatase and tensin homolog (PTEN) is a vital protein that maintains an inhibitory brake for cellular proliferation and growth. Accordingly, PTEN loss-of-function mutations are associated with a broad spectrum of human pathologies. Despite its importance, there is currently no method to directly monitor PTEN activity with cellular specificity within intact biological systems. Here we describe the development of a FRET-based biosensor using PTEN conformation as a proxy for the PTEN activity state, for two-photon fluorescence lifetime imaging microscopy. We identify a point mutation that allows the monitoring of PTEN activity with minimal interference to endogenous PTEN signaling. We demonstrate imaging of PTEN activity in cell lines, intact Caenorhabditis elegans and in the mouse brain. Finally, we develop a red-shifted sensor variant that allows us to identify cell-type-specific PTEN activity in excitatory and inhibitory cortical cells. In summary, our approach enables dynamic imaging of PTEN activity in vivo with unprecedented spatial and temporal resolution.</p>","PeriodicalId":18981,"journal":{"name":"Nature Methods","volume":" ","pages":""},"PeriodicalIF":36.1,"publicationDate":"2025-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143468633","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-18DOI: 10.1038/s41592-025-02594-6
Amine Driouchi, Mason Bretan, Brynmor J. Davis, Alec Heckert, Markus Seeger, Maité Bradley Silva, William S. R. Forrest, Jessica Hsiung, Jiongyi Tan, Hongli Yang, David T. McSwiggen, Linda Song, Askhay Sule, Behnam Abaie, Hanzhe Chen, Bryant Chhun, Brianna Conroy, Liam A. Elliott, Eric Gonzalez, Fedor Ilkov, Joshua Isaacs, George Labaria, Michelle Lagana, DeLaine D. Larsen, Brian Margolin, Mai K. Nguyen, Eugene Park, Jeremy Rine, Yangzhong Tang, Martin Vana, Andrew Wilkey, Zhengjian Zhang, Stephen Basham, Jaclyn J. Ho, Stephanie Johnson, Aaron A. Klammer, Kevin Lin, Xavier Darzacq, Eric Betzig, Russell T. Berman, Daniel J. Anderson
An ideal tool for the study of cellular biology would enable the measure of molecular activity nondestructively within living cells. Single-molecule localization microscopy (SMLM) techniques, such as single-molecule tracking (SMT), enable in situ measurements in cells but have historically been limited by a necessary tradeoff between spatiotemporal resolution and throughput. Here we address these limitations using oblique line scan (OLS), a robust single-objective light-sheet-based illumination and detection modality that achieves nanoscale spatial resolution and sub-millisecond temporal resolution across a large field of view. We show that OLS can be used to capture protein motion up to 14 μm2 s−1 in living cells. We further extend the utility of OLS with in-solution SMT for single-molecule measurement of ligand–protein interactions and disruption of protein–protein interactions using purified proteins. We illustrate the versatility of OLS by showcasing two-color SMT, STORM and single-molecule fluorescence recovery after photobleaching. OLS paves the way for robust, high-throughput, single-molecule investigations of protein function required for basic research, drug screening and systems biology studies. Oblique line scan microscopy achieves nanoscale spatial and sub-millisecond temporal resolution across a large field of view, enabling improved and robust single-molecule biophysical measurements and single-molecule tracking in both cells and solution.
{"title":"Oblique line scan illumination enables expansive, accurate and sensitive single-protein measurements in solution and in living cells","authors":"Amine Driouchi, Mason Bretan, Brynmor J. Davis, Alec Heckert, Markus Seeger, Maité Bradley Silva, William S. R. Forrest, Jessica Hsiung, Jiongyi Tan, Hongli Yang, David T. McSwiggen, Linda Song, Askhay Sule, Behnam Abaie, Hanzhe Chen, Bryant Chhun, Brianna Conroy, Liam A. Elliott, Eric Gonzalez, Fedor Ilkov, Joshua Isaacs, George Labaria, Michelle Lagana, DeLaine D. Larsen, Brian Margolin, Mai K. Nguyen, Eugene Park, Jeremy Rine, Yangzhong Tang, Martin Vana, Andrew Wilkey, Zhengjian Zhang, Stephen Basham, Jaclyn J. Ho, Stephanie Johnson, Aaron A. Klammer, Kevin Lin, Xavier Darzacq, Eric Betzig, Russell T. Berman, Daniel J. Anderson","doi":"10.1038/s41592-025-02594-6","DOIUrl":"10.1038/s41592-025-02594-6","url":null,"abstract":"An ideal tool for the study of cellular biology would enable the measure of molecular activity nondestructively within living cells. Single-molecule localization microscopy (SMLM) techniques, such as single-molecule tracking (SMT), enable in situ measurements in cells but have historically been limited by a necessary tradeoff between spatiotemporal resolution and throughput. Here we address these limitations using oblique line scan (OLS), a robust single-objective light-sheet-based illumination and detection modality that achieves nanoscale spatial resolution and sub-millisecond temporal resolution across a large field of view. We show that OLS can be used to capture protein motion up to 14 μm2 s−1 in living cells. We further extend the utility of OLS with in-solution SMT for single-molecule measurement of ligand–protein interactions and disruption of protein–protein interactions using purified proteins. We illustrate the versatility of OLS by showcasing two-color SMT, STORM and single-molecule fluorescence recovery after photobleaching. OLS paves the way for robust, high-throughput, single-molecule investigations of protein function required for basic research, drug screening and systems biology studies. Oblique line scan microscopy achieves nanoscale spatial and sub-millisecond temporal resolution across a large field of view, enabling improved and robust single-molecule biophysical measurements and single-molecule tracking in both cells and solution.","PeriodicalId":18981,"journal":{"name":"Nature Methods","volume":"22 3","pages":"559-568"},"PeriodicalIF":36.1,"publicationDate":"2025-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.com/articles/s41592-025-02594-6.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143449582","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-17DOI: 10.1038/s41592-024-02582-2
Adam Gosztolai, Robert L. Peach, Alexis Arnaudon, Mauricio Barahona, Pierre Vandergheynst
The dynamics of neuron populations commonly evolve on low-dimensional manifolds. Thus, we need methods that learn the dynamical processes over neural manifolds to infer interpretable and consistent latent representations. We introduce a representation learning method, MARBLE, which decomposes on-manifold dynamics into local flow fields and maps them into a common latent space using unsupervised geometric deep learning. In simulated nonlinear dynamical systems, recurrent neural networks and experimental single-neuron recordings from primates and rodents, we discover emergent low-dimensional latent representations that parametrize high-dimensional neural dynamics during gain modulation, decision-making and changes in the internal state. These representations are consistent across neural networks and animals, enabling the robust comparison of cognitive computations. Extensive benchmarking demonstrates state-of-the-art within- and across-animal decoding accuracy of MARBLE compared to current representation learning approaches, with minimal user input. Our results suggest that a manifold structure provides a powerful inductive bias to develop decoding algorithms and assimilate data across experiments. MARBLE uses geometric deep learning to map dynamics such as neural activity into a latent representation, which can then be used to decode the neural activity or compare it across systems.
{"title":"MARBLE: interpretable representations of neural population dynamics using geometric deep learning","authors":"Adam Gosztolai, Robert L. Peach, Alexis Arnaudon, Mauricio Barahona, Pierre Vandergheynst","doi":"10.1038/s41592-024-02582-2","DOIUrl":"10.1038/s41592-024-02582-2","url":null,"abstract":"The dynamics of neuron populations commonly evolve on low-dimensional manifolds. Thus, we need methods that learn the dynamical processes over neural manifolds to infer interpretable and consistent latent representations. We introduce a representation learning method, MARBLE, which decomposes on-manifold dynamics into local flow fields and maps them into a common latent space using unsupervised geometric deep learning. In simulated nonlinear dynamical systems, recurrent neural networks and experimental single-neuron recordings from primates and rodents, we discover emergent low-dimensional latent representations that parametrize high-dimensional neural dynamics during gain modulation, decision-making and changes in the internal state. These representations are consistent across neural networks and animals, enabling the robust comparison of cognitive computations. Extensive benchmarking demonstrates state-of-the-art within- and across-animal decoding accuracy of MARBLE compared to current representation learning approaches, with minimal user input. Our results suggest that a manifold structure provides a powerful inductive bias to develop decoding algorithms and assimilate data across experiments. MARBLE uses geometric deep learning to map dynamics such as neural activity into a latent representation, which can then be used to decode the neural activity or compare it across systems.","PeriodicalId":18981,"journal":{"name":"Nature Methods","volume":"22 3","pages":"612-620"},"PeriodicalIF":36.1,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.com/articles/s41592-024-02582-2.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143441476","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-17DOI: 10.1038/s41592-025-02602-9
Vivien Marx
These researchers put their labs’ philosophies into practice and find it empowers science and collaboration.
{"title":"The value of lab values","authors":"Vivien Marx","doi":"10.1038/s41592-025-02602-9","DOIUrl":"10.1038/s41592-025-02602-9","url":null,"abstract":"These researchers put their labs’ philosophies into practice and find it empowers science and collaboration.","PeriodicalId":18981,"journal":{"name":"Nature Methods","volume":"22 3","pages":"451-451"},"PeriodicalIF":36.1,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143441479","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-17DOI: 10.1038/s41592-024-02581-3
The MARBLE method addresses a critical challenge in neural population recordings: inferring expressive and interpretable latent representations that are comparable across experiments and animals. It achieves this by explicitly leveraging the low-dimensional structure of neural states through geometric deep learning to learn the dynamical flow fields in neural activity.
{"title":"Interpreting and comparing neural activity across systems by geometric deep learning","authors":"","doi":"10.1038/s41592-024-02581-3","DOIUrl":"10.1038/s41592-024-02581-3","url":null,"abstract":"The MARBLE method addresses a critical challenge in neural population recordings: inferring expressive and interpretable latent representations that are comparable across experiments and animals. It achieves this by explicitly leveraging the low-dimensional structure of neural states through geometric deep learning to learn the dynamical flow fields in neural activity.","PeriodicalId":18981,"journal":{"name":"Nature Methods","volume":"22 3","pages":"467-468"},"PeriodicalIF":36.1,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143441475","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-17DOI: 10.1038/s41592-025-02609-2
Vivien Marx
Vectors can be the ultimate vehicle for transporting material into cells. Except when they’re not so ultimate.
{"title":"Vector choices, vector surprises","authors":"Vivien Marx","doi":"10.1038/s41592-025-02609-2","DOIUrl":"10.1038/s41592-025-02609-2","url":null,"abstract":"Vectors can be the ultimate vehicle for transporting material into cells. Except when they’re not so ultimate.","PeriodicalId":18981,"journal":{"name":"Nature Methods","volume":"22 3","pages":"454-458"},"PeriodicalIF":36.1,"publicationDate":"2025-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.com/articles/s41592-025-02609-2.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143441528","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-02-12DOI: 10.1038/s41592-024-02580-4
Anwai Archit, Luca Freckmann, Sushmita Nair, Nabeel Khalid, Paul Hilt, Vikas Rajashekar, Marei Freitag, Carolin Teuber, Genevieve Buckley, Sebastian von Haaren, Sagnik Gupta, Andreas Dengel, Sheraz Ahmed, Constantin Pape
Accurate segmentation of objects in microscopy images remains a bottleneck for many researchers despite the number of tools developed for this purpose. Here, we present Segment Anything for Microscopy (μSAM), a tool for segmentation and tracking in multidimensional microscopy data. It is based on Segment Anything, a vision foundation model for image segmentation. We extend it by fine-tuning generalist models for light and electron microscopy that clearly improve segmentation quality for a wide range of imaging conditions. We also implement interactive and automatic segmentation in a napari plugin that can speed up diverse segmentation tasks and provides a unified solution for microscopy annotation across different microscopy modalities. Our work constitutes the application of vision foundation models in microscopy, laying the groundwork for solving image analysis tasks in this domain with a small set of powerful deep learning models. Segment Anything for Microscopy (μSAM) builds on the vision foundation model Segment Anything for high-quality image segmentation over a wide range of imaging conditions including light and electron microscopy.
{"title":"Segment Anything for Microscopy","authors":"Anwai Archit, Luca Freckmann, Sushmita Nair, Nabeel Khalid, Paul Hilt, Vikas Rajashekar, Marei Freitag, Carolin Teuber, Genevieve Buckley, Sebastian von Haaren, Sagnik Gupta, Andreas Dengel, Sheraz Ahmed, Constantin Pape","doi":"10.1038/s41592-024-02580-4","DOIUrl":"10.1038/s41592-024-02580-4","url":null,"abstract":"Accurate segmentation of objects in microscopy images remains a bottleneck for many researchers despite the number of tools developed for this purpose. Here, we present Segment Anything for Microscopy (μSAM), a tool for segmentation and tracking in multidimensional microscopy data. It is based on Segment Anything, a vision foundation model for image segmentation. We extend it by fine-tuning generalist models for light and electron microscopy that clearly improve segmentation quality for a wide range of imaging conditions. We also implement interactive and automatic segmentation in a napari plugin that can speed up diverse segmentation tasks and provides a unified solution for microscopy annotation across different microscopy modalities. Our work constitutes the application of vision foundation models in microscopy, laying the groundwork for solving image analysis tasks in this domain with a small set of powerful deep learning models. Segment Anything for Microscopy (μSAM) builds on the vision foundation model Segment Anything for high-quality image segmentation over a wide range of imaging conditions including light and electron microscopy.","PeriodicalId":18981,"journal":{"name":"Nature Methods","volume":"22 3","pages":"579-591"},"PeriodicalIF":36.1,"publicationDate":"2025-02-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.nature.com/articles/s41592-024-02580-4.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143409309","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号