{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,25]],"date-time":"2026-09-25T13:12:40Z","timestamp":1790341960253,"version":"4.1.0"},"reference-count":24,"publisher":"BMJ","issue":"7","funder":[{"name":"the National Key Research and Development Program of China","award":["2019YFC0118400"],"award-info":[{"award-number":["2019YFC0118400"]}]},{"name":"the Clinical Medical Research Center for Eye Diseases of Zhejiang Province","award":["2021E50007"],"award-info":[{"award-number":["2021E50007"]}]}],"content-domain":{"domain":["bmj.com"],"crossmark-restriction":true},"short-container-title":["Br J Ophthalmol"],"accepted":{"date-parts":[[2023,9,3]]},"published-print":{"date-parts":[[2024,7]]},"abstract":"<jats:sec>\n<jats:title>Background<\/jats:title>\n<jats:p>Ultrasound imaging is suitable for detecting and diagnosing ophthalmic abnormalities. However, a shortage of experienced sonographers and ophthalmologists remains a problem. This study aims to develop a multibranch transformer network (MBT-Net) for the automated classification of multiple ophthalmic diseases using B-mode ultrasound images.<\/jats:p>\n<\/jats:sec>\n<jats:sec>\n<jats:title>Methods<\/jats:title>\n<jats:p>Ultrasound images with six clinically confirmed categories, including normal, retinal detachment, vitreous haemorrhage, intraocular tumour, posterior scleral staphyloma and other abnormalities, were used to develop and evaluate the MBT-Net. Images were derived from five different ultrasonic devices operated by different sonographers and divided into training set, validation set, internal testing set and temporal external testing set. Two senior ophthalmologists and two junior ophthalmologists were recruited to compare the model\u2019s performance.<\/jats:p>\n<\/jats:sec>\n<jats:sec>\n<jats:title>Results<\/jats:title>\n    <jats:p>A total of 10\u2009184 ultrasound images were collected. The MBT-Net got an accuracy of 87.80% (95% CI 86.26% to 89.18%) in the internal testing set, which was significantly higher than junior ophthalmologists (95% CI 67.37% to 79.16%; both p&lt;0.05) and lower than senior ophthalmologists (95% CI 89.45% to 92.61%; both p&lt;0.05). The micro-average area under the curve of the six-category classification was 0.98. With reference to comprehensive clinical diagnosis, the measurements of agreement were almost perfect in the MBT-Net (kappa=0.85, p&lt;0.05). There was no significant difference in the accuracy of the MBT-Net across five ultrasonic devices (p=0.27). The MBT-Net got an accuracy of 82.21% (95% CI 78.45% to 85.44%) in the temporal external testing set.<\/jats:p>\n<\/jats:sec>\n<jats:sec>\n<jats:title>Conclusions<\/jats:title>\n<jats:p>The MBT-Net showed high accuracy for screening and diagnosing multiple ophthalmic diseases using only ultrasound images across mutioperators and mutidevices.<\/jats:p>\n<\/jats:sec>","DOI":"10.1136\/bjo-2022-322953","type":"journal-article","created":{"date-parts":[[2023,10,18]],"date-time":"2023-10-18T17:20:20Z","timestamp":1697649620000},"page":"999-1004","update-policy":"https:\/\/doi.org\/10.1136\/crossmarkpolicy","source":"Crossref","is-referenced-by-count":9,"title":["Automated classification of multiple ophthalmic diseases using ultrasound images by deep learning"],"prefix":"10.1136","volume":"108","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-9454-8952","authenticated-orcid":false,"given":"Yijie","family":"Wang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zihao","family":"Xu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7800-9297","authenticated-orcid":false,"given":"Ruilong","family":"Dan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chunlei","family":"Yao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ji","family":"Shao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yiming","family":"Sun","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yaqi","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1948-2500","authenticated-orcid":false,"given":"Juan","family":"Ye","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"239","published-online":{"date-parts":[[2023,10,18]]},"reference":[{"key":"2024062002150809000_108.7.999.1","doi-asserted-by":"crossref","first-page":"100759","DOI":"10.1016\/j.preteyeres.2019.04.003","article-title":"Deep learning in ophthalmology: the technical and clinical considerations","volume":"72","author":"Ting","year":"2019","journal-title":"Prog Retin Eye Res"},{"key":"2024062002150809000_108.7.999.2","doi-asserted-by":"publisher","DOI":"10.1001\/jama.2016.17216"},{"key":"2024062002150809000_108.7.999.3","doi-asserted-by":"publisher","DOI":"10.1136\/bjophthalmol-2018-313173"},{"key":"2024062002150809000_108.7.999.4","doi-asserted-by":"crossref","first-page":"351","DOI":"10.1007\/s13244-016-0471-z","article-title":"Ocular ultrasonography focused on the posterior eye segment: what radiologists should know","volume":"7","author":"De La Hoz Polo","year":"2016","journal-title":"Insights Imaging"},{"key":"2024062002150809000_108.7.999.5","doi-asserted-by":"publisher","DOI":"10.1177\/014107688007300411"},{"key":"2024062002150809000_108.7.999.6","doi-asserted-by":"crossref","first-page":"109717","DOI":"10.1016\/j.ejrad.2021.109717","article-title":"Artificial intelligence in ultrasound","volume":"139","author":"Shen","year":"2021","journal-title":"Eur J Radiol"},{"key":"2024062002150809000_108.7.999.7","doi-asserted-by":"crossref","first-page":"416","DOI":"10.1007\/s12072-019-09937-4","article-title":"Current status and perspectives for computer-aided ultrasonic diagnosis of liver lesions using deep learning technology","volume":"13","author":"Nishida","year":"2019","journal-title":"Hepatol Int"},{"key":"2024062002150809000_108.7.999.8","doi-asserted-by":"crossref","DOI":"10.3390\/diagnostics10121055","article-title":"The utility of deep learning in breast ultrasonic imaging: a review","volume":"10","author":"Fujioka","year":"2020","journal-title":"Diagnostics (Basel)"},{"key":"2024062002150809000_108.7.999.9","doi-asserted-by":"crossref","DOI":"10.3390\/cancers13194740","article-title":"Artificial intelligence in thyroid field-a comprehensive review","volume":"13","author":"Bini","year":"2021","journal-title":"Cancers (Basel)"},{"key":"2024062002150809000_108.7.999.10","doi-asserted-by":"crossref","first-page":"1684","DOI":"10.1109\/JBHI.2021.3129245","article-title":"AGMB-transformer: anatomy-guided multi-branch transformer network for automated evaluation of root canal therapy","volume":"26","author":"Li","year":"2022","journal-title":"IEEE J Biomed Health Inform"},{"key":"2024062002150809000_108.7.999.11","doi-asserted-by":"crossref","first-page":"193","DOI":"10.1007\/BF03178082","article-title":"Contrast limited adaptive Histogram equalization image processing to improve the detection of simulated spiculations in dense mammograms","volume":"11","author":"Pisano","year":"1998","journal-title":"J Digit Imaging"},{"key":"2024062002150809000_108.7.999.12","unstructured":"Diederik P , Kingma JLB . Adam: a method for stochastic optimization. International Conference on Learning Representations (ICLR); 2015"},{"key":"2024062002150809000_108.7.999.13","doi-asserted-by":"publisher","DOI":"10.1177\/001316446002000104"},{"key":"2024062002150809000_108.7.999.14","doi-asserted-by":"crossref","unstructured":"Nelson BP , Topol E , Bhagra A , et al . Atlas of handheld ultrasound.1st ed. Cham: Springer, 2018. doi:10.1007\/978-3-319-73855-0","DOI":"10.1007\/978-3-319-73855-0"},{"key":"2024062002150809000_108.7.999.15","doi-asserted-by":"crossref","first-page":"100021","DOI":"10.1016\/j.aopr.2021.100021","article-title":"The barriers to global eye care equity and the role of digital innovations","volume":"1","author":"Zheng","year":"2021","journal-title":"Advances in Ophthalmology Practice and Research"},{"key":"2024062002150809000_108.7.999.16","doi-asserted-by":"crossref","first-page":"101555","DOI":"10.1016\/j.media.2019.101555","article-title":"Automated detection and classification of thyroid nodules in ultrasound images using clinical-knowledge-guided convolutional neural networks","volume":"58","author":"Liu","year":"2019","journal-title":"Med Image Anal"},{"key":"2024062002150809000_108.7.999.17","doi-asserted-by":"crossref","first-page":"1264","DOI":"10.1007\/s00330-019-06407-1","article-title":"Deep learning with ultrasonography: automated classification of liver fibrosis using a deep convolutional neural network","volume":"30","author":"Lee","year":"2020","journal-title":"Eur Radiol"},{"key":"2024062002150809000_108.7.999.18","doi-asserted-by":"crossref","first-page":"1145","DOI":"10.4103\/ijo.IJO_2119_21","article-title":"Development and validation of an offline deep learning algorithm to detect vitreoretinal abnormalities on ocular ultrasound","volume":"70","author":"Adithya","year":"2022","journal-title":"Indian J Ophthalmol"},{"key":"2024062002150809000_108.7.999.19","doi-asserted-by":"crossref","DOI":"10.1167\/tvst.10.4.22","article-title":"A deep learning model for screening multiple abnormal findings in ophthalmic ultrasonography (with video)","volume":"10","author":"Chen","year":"2021","journal-title":"Transl Vis Sci Technol"},{"key":"2024062002150809000_108.7.999.20","unstructured":"Simonyan K , Zisserman A . Very deep convolutional networks for large-scale image recognition. 2015 International Conference on Learning Representations (ICLR); 2015"},{"key":"2024062002150809000_108.7.999.21","doi-asserted-by":"crossref","unstructured":"He K , Zhang X , Ren S , et al . Deep residual learning for image recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR); 2016","DOI":"10.1109\/CVPR.2016.90"},{"key":"2024062002150809000_108.7.999.22","doi-asserted-by":"crossref","unstructured":"Szegedy C , Vanhoucke V , Ioffe S , et al . Rethinking the inception architecture for computer vision. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR); 2016","DOI":"10.1109\/CVPR.2016.308"},{"key":"2024062002150809000_108.7.999.23","doi-asserted-by":"crossref","first-page":"87","DOI":"10.1109\/TPAMI.2022.3152247","article-title":"A survey on vision transformer","volume":"45","author":"Han","year":"2023","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"2024062002150809000_108.7.999.24","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2023.102802"}],"container-title":["British Journal of Ophthalmology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1136\/bjo-2022-322953","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,6,20]],"date-time":"2024-06-20T18:29:50Z","timestamp":1718908190000},"score":1,"resource":{"primary":{"URL":"https:\/\/bjo.bmj.com\/lookup\/doi\/10.1136\/bjo-2022-322953"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,10,18]]},"references-count":24,"journal-issue":{"issue":"7","published-online":{"date-parts":[[2024,6,20]]},"published-print":{"date-parts":[[2024,7]]}},"alternative-id":["10.1136\/bjo-2022-322953"],"URL":"https:\/\/doi.org\/10.1136\/bjo-2022-322953","relation":{},"ISSN":["0007-1161","1468-2079"],"issn-type":[{"value":"0007-1161","type":"print"},{"value":"1468-2079","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,10,18]]}}}