{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,28]],"date-time":"2026-02-28T02:06:52Z","timestamp":1772244412947,"version":"3.50.1"},"reference-count":27,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2023,12,15]],"date-time":"2023-12-15T00:00:00Z","timestamp":1702598400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Bioinform."],"abstract":"<jats:p>Electron microscopy (EM) enables imaging at a resolution of nanometers and can shed light on how cancer evolves to develop resistance to therapy. Acquiring these images has become a routine task.However, analyzing them is now a bottleneck, as manual structure identification is very time-consuming and can take up to several months for a single sample. Deep learning approaches offer a suitable solution to speed up the analysis. In this work, we present a study of several state-of-the-art deep learning models for the task of segmenting nuclei and nucleoli in volumes from tumor biopsies. We compared previous results obtained with the ResUNet architecture to the more recent UNet++, FracTALResNet, SenFormer, and CEECNet models. In addition, we explored the utilization of unlabeled images through semi-supervised learning with Cross Pseudo Supervision. We have trained and evaluated all of the models on sparse manual labels from three fully annotated in-house datasets that we have made available on demand, demonstrating improvements in terms of 3D Dice score. From the analysis of these results, we drew conclusions on the relative gains of using more complex models, and semi-supervised learning as well as the next steps for the mitigation of the manual segmentation bottleneck.<\/jats:p>","DOI":"10.3389\/fbinf.2023.1308707","type":"journal-article","created":{"date-parts":[[2023,12,15]],"date-time":"2023-12-15T03:54:18Z","timestamp":1702612458000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Efficient semi-supervised semantic segmentation of electron microscopy cancer images with sparse annotations"],"prefix":"10.3389","volume":"3","author":[{"given":"Lucas","family":"Pagano","sequence":"first","affiliation":[]},{"given":"Guillaume","family":"Thibault","sequence":"additional","affiliation":[]},{"given":"Walid","family":"Bousselham","sequence":"additional","affiliation":[]},{"given":"Jessica L.","family":"Riesterer","sequence":"additional","affiliation":[]},{"given":"Xubo","family":"Song","sequence":"additional","affiliation":[]},{"given":"Joe W.","family":"Gray","sequence":"additional","affiliation":[]}],"member":"1965","published-online":{"date-parts":[[2023,12,15]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"59","DOI":"10.1186\/s12964-020-0530-4","article-title":"Tumor microenvironment complexity and therapeutic implications at a glance","volume":"18","author":"Baghban","year":"2020","journal-title":"Cell. Commun. Signal"},{"key":"B2","article-title":"Efficient self-ensemble for semantic segmentation","author":"Bousselham","year":"2022"},{"key":"B3","doi-asserted-by":"publisher","DOI":"10.1038\/nprot.2011.332","article-title":"Imaging three-dimensional tissue architectures by focused ion beam scanning electron microscopy","author":"Bushby","year":"2011"},{"key":"B4","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR46437.2021.00264","article-title":"Semi-supervised semantic segmentation with cross pseudo supervision","author":"Chen","year":"2021"},{"key":"B5","doi-asserted-by":"publisher","first-page":"3707","DOI":"10.3390\/RS13183707","article-title":"Looking for change? roll the dice and demand attention","volume":"2021","author":"Diakogiannis","year":"2021","journal-title":"Remote Sens."},{"key":"B6","doi-asserted-by":"crossref","DOI":"10.1007\/b101190","volume-title":"Introduction to focused ion beams: instrumentation, theory, techniques and practice","author":"Giannuzzi","year":"2005"},{"key":"B7","first-page":"249","article-title":"Understanding the difficulty of training deep feedforward neural networks","volume":"9","author":"Glorot","year":"2010","journal-title":"J. Mach. Learn. Res. - Proc. Track"},{"key":"B8","first-page":"770","article-title":"Deep residual learning for image recognition","author":"He","year":"2016"},{"key":"B9","doi-asserted-by":"publisher","first-page":"a026781","DOI":"10.1101\/cshperspect.a026781","article-title":"Tumor microenvironment and differential responses to Therapy","volume":"7","author":"Hirata","year":"2017","journal-title":"Cold Spring Harb. Perspect. Med."},{"key":"B10","article-title":"Semask: semantically masked transformers for semantic segmentation","author":"Jain","year":"2021"},{"key":"B11","doi-asserted-by":"publisher","first-page":"e0230605","DOI":"10.1371\/JOURNAL.PONE.0230605","article-title":"Semantic segmentation of hela cells: an objective comparison between one traditional algorithm and four deep-learning architectures","volume":"15","author":"Karaba\u01e7","year":"2020","journal-title":"PLoS ONE"},{"key":"B12","article-title":"A formaldehyde-glutaraldehyde fixative of high osmolality for use in electron microscopy","volume":"27","author":"Karnovsky","year":"1964","journal-title":"J. Cell. Biol."},{"key":"B13","doi-asserted-by":"publisher","first-page":"2351","DOI":"10.1038\/s41388-017-0121-z","article-title":"Nucleolus as an emerging hub in maintenance of genome stability and cancer pathogenesis","volume":"37","author":"Lindstr\u00f6m","year":"2018","journal-title":"Oncogene"},{"key":"B14","doi-asserted-by":"publisher","DOI":"10.1101\/2021.05.27.446019","article-title":"Robust segmentation of cellular ultrastructure on sparsely labeled 3d electron microscopy images using deep learning","author":"Machireddy","year":"2021","journal-title":"bioRxiv"},{"key":"B15","doi-asserted-by":"publisher","first-page":"20140177","DOI":"10.1098\/rstb.2014.0177","article-title":"Peto\u2019s paradox and the promise of comparative oncology","volume":"370","author":"Nunney","year":"2015","journal-title":"Philosophical Trans. R. Soc. B Biol. Sci."},{"key":"B16","doi-asserted-by":"publisher","first-page":"126","DOI":"10.3389\/fnana.2014.00126","article-title":"A workflow for the automatic segmentation of organelles in electron microscopy image stacks","volume":"8","author":"Perez","year":"2014","journal-title":"Front. Neuroanat."},{"key":"B17","doi-asserted-by":"publisher","DOI":"10.1101\/675371","article-title":"A workflow for visualizing human cancer biopsies using large-format electron microscopy","author":"Riesterer","year":"2019","journal-title":"bioRxiv"},{"key":"B18","doi-asserted-by":"publisher","first-page":"20","DOI":"10.1038\/nrc.2016.108","article-title":"Cancer nanomedicine: progress, challenges and opportunities","volume":"17","author":"Shi","year":"2017","journal-title":"Nat. Rev. Cancer"},{"key":"B19","doi-asserted-by":"publisher","first-page":"82031","DOI":"10.1109\/access.2021.3086020","article-title":"U-net and its variants for medical image segmentation: a review of theory and applications","volume":"9","author":"Siddique","year":"2020","journal-title":"IEEE Access"},{"key":"B20","doi-asserted-by":"publisher","first-page":"639930","DOI":"10.3389\/fgene.2021.639930","article-title":"Msu-net: multi-scale u-net for 2d medical image segmentation","volume":"12","author":"Su","year":"2021","journal-title":"Front. Genet."},{"key":"B21","doi-asserted-by":"publisher","first-page":"29","DOI":"10.1186\/s12880-015-0068-x","article-title":"Metrics for evaluating 3d medical image segmentation: analysis, selection, and tool","volume":"15","author":"Taha","year":"2015","journal-title":"BMC Med. Imaging"},{"key":"B22","doi-asserted-by":"publisher","first-page":"2085\u2013","DOI":"10.1111\/cas.13630","article-title":"Stromal barriers to nanomedicine penetration in the pancreatic tumor microenvironment","volume":"109","author":"Tanaka","year":"2018","journal-title":"Cancer Sci."},{"key":"B23","doi-asserted-by":"publisher","first-page":"27","DOI":"10.1109\/83.650848","article-title":"A pyramid approach to subpixel registration based on intensity","volume":"7","author":"Thevenaz","year":"1998","journal-title":"IEEE Trans. Image Process."},{"key":"B24","first-page":"6022","article-title":"Cutmix: regularization strategy to train strong classifiers with localizable features","author":"Yun","year":"2019"},{"key":"B25","doi-asserted-by":"publisher","first-page":"117934","DOI":"10.1016\/J.NEUROIMAGE.2021.117934","article-title":"Deep learning based segmentation of brain tissue from diffusion mri","volume":"233","author":"Zhang","year":"2021","journal-title":"NeuroImage"},{"key":"B26","doi-asserted-by":"publisher","first-page":"10165","DOI":"10.1007\/978-3-030-00889-5_1","article-title":"Unet++: a nested u-net architecture for medical image segmentation","author":"Zhou","year":"2018","journal-title":"Corr. abs\/1807"},{"key":"B27","doi-asserted-by":"publisher","first-page":"677","DOI":"10.1038\/NRC1430","article-title":"Nuclear structure in cancer cells","volume":"4","author":"Zink","year":"2004","journal-title":"Nat. Rev. Cancer"}],"container-title":["Frontiers in Bioinformatics"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fbinf.2023.1308707\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,12,15]],"date-time":"2023-12-15T03:54:22Z","timestamp":1702612462000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fbinf.2023.1308707\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,12,15]]},"references-count":27,"alternative-id":["10.3389\/fbinf.2023.1308707"],"URL":"https:\/\/doi.org\/10.3389\/fbinf.2023.1308707","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2023.10.30.563998","asserted-by":"object"}]},"ISSN":["2673-7647"],"issn-type":[{"value":"2673-7647","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,12,15]]},"article-number":"1308707"}}