{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,28]],"date-time":"2026-02-28T03:23:12Z","timestamp":1772248992490,"version":"3.50.1"},"reference-count":36,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2024,5,9]],"date-time":"2024-05-09T00:00:00Z","timestamp":1715212800000},"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. Comput. Neurosci."],"abstract":"<jats:p>\n                    Automatic segmentation of vestibular schwannoma (VS) from routine clinical MRI has potential to improve clinical workflow, facilitate treatment decisions, and assist patient management. Previous work demonstrated reliable automatic segmentation performance on datasets of standardized MRI images acquired for stereotactic surgery planning. However, diagnostic clinical datasets are generally more diverse and pose a larger challenge to automatic segmentation algorithms, especially when post-operative images are included. In this work, we show for the first time that automatic segmentation of VS on routine MRI datasets is also possible with high accuracy. We acquired and publicly release a curated multi-center routine clinical (MC-RC) dataset of 160 patients with a single sporadic VS. For each patient up to three longitudinal MRI exams with contrast-enhanced T1-weighted (ceT1w) (\n                    <jats:italic>n<\/jats:italic>\n                    = 124) and T2-weighted (T2w) (\n                    <jats:italic>n<\/jats:italic>\n                    = 363) images were included and the VS manually annotated. Segmentations were produced and verified in an iterative process: (1) initial segmentations by a specialized company; (2) review by one of three trained radiologists; and (3) validation by an expert team. Inter- and intra-observer reliability experiments were performed on a subset of the dataset. A state-of-the-art deep learning framework was used to train segmentation models for VS. Model performance was evaluated on a MC-RC hold-out testing set, another public VS datasets, and a partially public dataset. The generalizability and robustness of the VS deep learning segmentation models increased significantly when trained on the MC-RC dataset. Dice similarity coefficients (DSC) achieved by our model are comparable to those achieved by trained radiologists in the inter-observer experiment. On the MC-RC testing set, median DSCs were 86.2(9.5) for ceT1w, 89.4(7.0) for T2w, and 86.4(8.6) for combined ceT1w+T2w input images. On another public dataset acquired for Gamma Knife stereotactic radiosurgery our model achieved median DSCs of 95.3(2.9), 92.8(3.8), and 95.5(3.3), respectively. In contrast, models trained on the Gamma Knife dataset did not generalize well as illustrated by significant underperformance on the MC-RC routine MRI dataset, highlighting the importance of data variability in the development of robust VS segmentation models. The MC-RC dataset and all trained deep learning models were made available online.\n                  <\/jats:p>","DOI":"10.3389\/fncom.2024.1365727","type":"journal-article","created":{"date-parts":[[2024,5,9]],"date-time":"2024-05-09T01:00:59Z","timestamp":1715216459000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":12,"title":["Deep learning for automatic segmentation of vestibular schwannoma: a retrospective study from multi-center routine MRI"],"prefix":"10.3389","volume":"18","author":[{"given":"Aaron","family":"Kujawa","sequence":"first","affiliation":[]},{"given":"Reuben","family":"Dorent","sequence":"additional","affiliation":[]},{"given":"Steve","family":"Connor","sequence":"additional","affiliation":[]},{"given":"Suki","family":"Thomson","sequence":"additional","affiliation":[]},{"given":"Marina","family":"Ivory","sequence":"additional","affiliation":[]},{"given":"Ali","family":"Vahedi","sequence":"additional","affiliation":[]},{"given":"Emily","family":"Guilhem","sequence":"additional","affiliation":[]},{"given":"Navodini","family":"Wijethilake","sequence":"additional","affiliation":[]},{"given":"Robert","family":"Bradford","sequence":"additional","affiliation":[]},{"given":"Neil","family":"Kitchen","sequence":"additional","affiliation":[]},{"given":"Sotirios","family":"Bisdas","sequence":"additional","affiliation":[]},{"given":"Sebastien","family":"Ourselin","sequence":"additional","affiliation":[]},{"given":"Tom","family":"Vercauteren","sequence":"additional","affiliation":[]},{"given":"Jonathan","family":"Shapey","sequence":"additional","affiliation":[]}],"member":"1965","published-online":{"date-parts":[[2024,5,9]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-022-30695-9","article-title":"The medical segmentation decathlon","author":"Antonelli","year":"2022","journal-title":"Nat. Commun"},{"key":"B2","doi-asserted-by":"publisher","first-page":"440","DOI":"10.1177\/0194599815590105","article-title":"The changing landscape of vestibular schwannoma management in the United States\u2013a shift toward conservatism","volume":"153","author":"Carlson","year":"2015","journal-title":"Otolaryngol. Head Neck Surg"},{"key":"B3","doi-asserted-by":"publisher","first-page":"834","DOI":"10.1109\/TPAMI.2017.2699184","article-title":"Deeplab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs","volume":"40","author":"Chen","year":"2017","journal-title":"IEEE Trans. Patt. Anal. Mach. Intell"},{"key":"B4","doi-asserted-by":"publisher","first-page":"1045","DOI":"10.1007\/s10278-013-9622-7","article-title":"The Cancer Imaging Archive (TCIA): maintaining and operating a public information repository","volume":"26","author":"Clark","year":"2013","journal-title":"J. Digit. Imag"},{"key":"B5","first-page":"647","article-title":"\u201cDecaf: a deep convolutional activation feature for generic visual recognition,\u201d","volume-title":"International Conference on Machine Learning","author":"Donahue","year":"2014"},{"key":"B6","doi-asserted-by":"publisher","first-page":"102628","DOI":"10.1016\/j.media.2022.102628","article-title":"CrossMoDA 2021 challenge: Benchmark of cross-modality domain adaptation techniques for vestibular schwannoma and cochlea segmentation","volume":"83","author":"Dorent","year":"2023","journal-title":"Med. Image Analy"},{"key":"B7","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-319-46976-8_19","article-title":"\u201cThe importance of skip connections in biomedical image segmentation,\u201d","volume-title":"International Workshop on Deep Learning in Medical Image Analysis, International Workshop on Large-Scale Annotation of Biomedical Data and Expert Label Synthesis","author":"Drozdzal","year":"2016"},{"key":"B8","doi-asserted-by":"publisher","first-page":"203","DOI":"10.1038\/s41592-020-01008-z","article-title":"nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation","volume":"18","author":"Isensee","year":"2021","journal-title":"Nat. Method"},{"key":"B9","doi-asserted-by":"publisher","first-page":"506","DOI":"10.3171\/jns.1998.88.3.0506","article-title":"Neurotopographic considerations in the microsurgical treatment of small acoustic neurinomas","volume":"88","author":"Koos","year":"1998","journal-title":"J. Neurosurg"},{"key":"B10","doi-asserted-by":"publisher","first-page":"837191","DOI":"10.3389\/fradi.2022.837191","article-title":"Automated Koos classification of vestibular schwannoma","volume":"2","author":"Kujawa","year":"2022","journal-title":"Front. Radiol"},{"key":"B11","doi-asserted-by":"publisher","DOI":"10.7937\/HRZH-2N82","article-title":"Segmentation of vestibular schwannoma from magnetic resonance imaging: an annotated multi-center routine clinical dataset (Vestibular-Schwannoma-MC-RC) (version 1) [dataset]","author":"Kujawa","year":"","journal-title":"The Cancer Imaging Archive"},{"key":"B12","doi-asserted-by":"publisher","DOI":"10.5281\/zenodo.10363647","article-title":"Deep learning for automatic segmentation of vestibular schwannoma: a retrospective study from multi-centre routine MRI \u2013 deep learning models","author":"Kujawa","year":"","journal-title":"Zenodo"},{"key":"B13","doi-asserted-by":"publisher","first-page":"3106","DOI":"10.1038\/s41598-021-82665-8","article-title":"Applying artificial intelligence to longitudinal imaging analysis of vestibular schwannoma following radiosurgery","volume":"11","author":"Lee","year":"2021","journal-title":"Sci. Rep"},{"key":"B14","doi-asserted-by":"publisher","first-page":"867","DOI":"10.1007\/s00405-018-4865-z","article-title":"A comparison of semi-automated volumetric vs linear measurement of small vestibular schwannomas","volume":"275","author":"MacKeith","year":"2018","journal-title":"Eur. Arch. Oto-Rhino-Laryngol"},{"key":"B15","article-title":"Metrics reloaded: pitfalls and recommendations for image analysis validation","author":"Maier-Hein","year":"2022","journal-title":"arXiv.org.2206.01653"},{"key":"B16","doi-asserted-by":"publisher","first-page":"1191","DOI":"10.1097\/MAO.0000000000001875","article-title":"Incidence of intralabyrinthine schwannoma: a population-based study within the United States","volume":"39","author":"Marinelli","year":"2018","journal-title":"Otol. Neurotol"},{"key":"B17","doi-asserted-by":"publisher","first-page":"024003","DOI":"10.1117\/1.JMI.1.2.024003","article-title":"Global image registration using a symmetric block-matching approach","volume":"1","author":"Modat","year":"2014","journal-title":"J. Med. Imag"},{"key":"B18","doi-asserted-by":"publisher","first-page":"e210300","DOI":"10.1148\/ryai.210300","article-title":"Fully automated 3D vestibular schwannoma segmentation with and without gadolinium-based contrast material: a multicenter, multivendor study","volume":"4","author":"Neve","year":"2022","journal-title":"Radiology"},{"key":"B19","doi-asserted-by":"publisher","first-page":"267","DOI":"10.1007\/s00701-006-1093-x","article-title":"Two-and three dimensional measures of vestibular schwannomas and posterior fossa-implications for the treatment","volume":"149","author":"Roche","year":"2007","journal-title":"Acta Neurochirur"},{"key":"B20","first-page":"234","article-title":"\u201cU-net: convolutional networks for biomedical image segmentation,\u201d","volume-title":"Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III","author":"Ronneberger","year":"2015"},{"key":"B21","doi-asserted-by":"publisher","first-page":"269","DOI":"10.1016\/j.wneu.2021.03.010","article-title":"Artificial intelligence opportunities for vestibular schwannoma management using image segmentation and clinical decision tools","volume":"149","author":"Shapey","year":"","journal-title":"World Neurosurg"},{"key":"B22","doi-asserted-by":"publisher","DOI":"10.7937\/TCIA.9YTJ-5Q73","article-title":"Segmentation of vestibular schwannoma from magnetic resonance imaging: an open annotated dataset and baseline algorithm (version 2) [Data set]","author":"Shapey","year":"","journal-title":"The Cancer Imaging Archive"},{"key":"B23","doi-asserted-by":"publisher","first-page":"286","DOI":"10.1038\/s41597-021-01064-w","article-title":"Segmentation of vestibular schwannoma from MRI, an open annotated dataset and baseline algorithm","volume":"8","author":"Shapey","year":"","journal-title":"Sci. Data"},{"key":"B24","doi-asserted-by":"publisher","first-page":"171","DOI":"10.3171\/2019.9.JNS191949","article-title":"An artificial intelligence framework for automatic segmentation and volumetry of vestibular schwannomas from contrast-enhanced T1-weighted and high-resolution T2-weighted MRI","volume":"134","author":"Shapey","year":"2019","journal-title":"J. Neurosurg"},{"key":"B25","doi-asserted-by":"publisher","first-page":"547","DOI":"10.1097\/00129492-200606000-00018","article-title":"The natural history of vestibular schwannoma","volume":"27","author":"Stangerup","year":"2006","journal-title":"Otol. Neurotol"},{"key":"B26","doi-asserted-by":"publisher","first-page":"1271","DOI":"10.1097\/MAO.0000000000000459","article-title":"Surveillance after resection of vestibular schwannoma: measurement techniques and predictors of growth","volume":"35","author":"Tang","year":"2014","journal-title":"Otol. Neurotol"},{"key":"B27","doi-asserted-by":"publisher","first-page":"1018","DOI":"10.1109\/TMI.2014.2366792","article-title":"Transfer learning improves supervised image segmentation across imaging protocols","volume":"34","author":"Van Opbroek","year":"2014","journal-title":"IEEE Trans. Med. Imag"},{"key":"B28","doi-asserted-by":"publisher","first-page":"706","DOI":"10.3171\/2011.12.JNS111662","article-title":"Growth of untreated vestibular schwannoma: a prospective study","volume":"116","author":"Varughese","year":"2012","journal-title":"J. Neurosurg"},{"key":"B29","doi-asserted-by":"publisher","first-page":"737","DOI":"10.1177\/0194599812447766","article-title":"Three-dimensional segmented volumetric analysis of sporadic vestibular schwannomas: comparison of segmented and linear measurements","volume":"147","author":"Walz","year":"2012","journal-title":"Otolaryngol.-Head Neck Surg"},{"key":"B30","doi-asserted-by":"publisher","first-page":"1562","DOI":"10.1109\/TMI.2018.2791721","article-title":"Interactive medical image segmentation using deep learning with image-specific fine tuning","volume":"37","author":"Wang","year":"","journal-title":"IEEE Trans. Med. Imag"},{"key":"B31","first-page":"264","article-title":"\u201cAutomatic segmentation of vestibular schwannoma from T2-weighted MRI by deep spatial attention with hardness-weighted loss,\u201d","volume-title":"Medical Image Computing and Computer Assisted Intervention-MICCAI 2019: 22nd International Conference, Shenzhen, China, October 13-17, 2019, Proceedings, Part II 22","author":"Wang","year":"2019"},{"key":"B32","doi-asserted-by":"publisher","first-page":"1559","DOI":"10.1109\/TPAMI.2018.2840695","article-title":"DeepIGeoS: a deep interactive geodesic framework for medical image segmentation","volume":"41","author":"Wang","year":"","journal-title":"IEEE Trans. Patt. Analy. Mach. Intell"},{"key":"B33","unstructured":"WijethilakeN.\n          CrossMoDA232023"},{"key":"B34","doi-asserted-by":"publisher","DOI":"10.2139\/ssrn.4664551","article-title":"Artificial intelligence for personalized management of vestibular schwannoma: a clinical implementation study within a multidisciplinary decision making environment","author":"Wijethilake","year":"2023","journal-title":"medRxiv"},{"key":"B35","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-17899-3_8","article-title":"Boundary distance loss for intra-\/extra-meatal segmentation of vestibular schwannoma","author":"Wijethilake","year":"2022","journal-title":"arXiv preprint arXiv:2208.04680"},{"key":"B36","doi-asserted-by":"publisher","first-page":"1116","DOI":"10.1016\/j.neuroimage.2006.01.015","article-title":"User-guided 3D active contour segmentation of anatomical structures: significantly improved efficiency and reliability","volume":"31","author":"Yushkevich","year":"2006","journal-title":"Neuroimage"}],"container-title":["Frontiers in Computational Neuroscience"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fncom.2024.1365727\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,5,9]],"date-time":"2024-05-09T01:01:11Z","timestamp":1715216471000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fncom.2024.1365727\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,5,9]]},"references-count":36,"alternative-id":["10.3389\/fncom.2024.1365727"],"URL":"https:\/\/doi.org\/10.3389\/fncom.2024.1365727","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2022.08.01.22278193","asserted-by":"object"}]},"ISSN":["1662-5188"],"issn-type":[{"value":"1662-5188","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,5,9]]},"article-number":"1365727"}}