{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,7]],"date-time":"2026-10-07T05:19:30Z","timestamp":1791350370121,"version":"4.3.1"},"reference-count":33,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2024,9,18]],"date-time":"2024-09-18T00:00:00Z","timestamp":1726617600000},"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. Med."],"abstract":"<jats:sec>\n                    <jats:title>Purpose<\/jats:title>\n                    <jats:p>This study aims to evaluate the diagnostic performance of a machine learning model (ML model) to train junior ophthalmologists in detecting preclinical keratoconus (PKC).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>A total of 1,334 corneal topography images (The Pentacam HR system) from 413 keratoconus eyes, 32 PKC eyes and 222 normal eyes were collected. Five junior ophthalmologists were trained and annotated the images with or without the suggestions proposed by the ML model. The diagnostic performance of PKC was evaluated among three groups: junior ophthalmologist group (control group), ML model group and ML model-training junior ophthalmologist group (test group).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      The accuracy of the ML model between the eyes of patients with KC and NEs in all three clinics (99% accuracy, area under the receiver operating characteristic (ROC) curve AUC of 1.00, 99% sensitivity, 99% specificity) was higher than that for Belin-Ambr\u00f3sio enhanced ectasia display total deviation (BAD-D) (86% accuracy, AUC of 0.97, 97% sensitivity, 69% specificity). The accuracy of the ML model between eyes with PKC and NEs in all three clinics (98% accuracy, AUC of 0.96, 98% sensitivity, 98% specificity) was higher than that of BAD-D (69% accuracy, AUC of 0.73, 67% sensitivity, 69% specificity). The diagnostic accuracy of PKC was 47.5% (95%CI, 0.5\u201371.6%), 100% (95%CI, 100\u2013100%) and 94.4% (95%CI, 14.7\u201394.7%) in the control group, ML model group and test group. With the assistance of the proposed ML model, the diagnostic accuracy of junior ophthalmologists improved with statistical significance (\n                      <jats:italic>p<\/jats:italic>\n                      \u2009&amp;lt;\u20090.05). According to the questionnaire of all the junior ophthalmologists, the average score was 4 (total 5) regarding to the comprehensiveness that the AI model has been in their keratoconus diagnosis learning; the average score was 4.4 (total 5) regarding to the convenience that the AI model has been in their keratoconus diagnosis learning.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>The proposed ML model provided a novel approach for the detection of PKC with high diagnostic accuracy and assisted to improve the performance of junior ophthalmologists, resulting especially in reducing the risk of missed diagnoses.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fmed.2024.1458356","type":"journal-article","created":{"date-parts":[[2024,9,18]],"date-time":"2024-09-18T01:04:25Z","timestamp":1726621465000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":4,"title":["The development of a machine learning model to train junior ophthalmologists in diagnosing the pre-clinical keratoconus"],"prefix":"10.3389","volume":"11","author":[{"given":"Yang","family":"Jiang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hanyu","family":"Jiang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jing","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tao","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ying","family":"Li","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yuehua","family":"Zhou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Youxin","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fusheng","family":"Li","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2024,9,18]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"502","DOI":"10.1016\/j.exer.2007.06.016","article-title":"Longitudinal study of keratoconus progression","volume":"85","author":"Li","year":"2007","journal-title":"Exp Eye Res"},{"key":"ref2","doi-asserted-by":"publisher","first-page":"515","DOI":"10.1016\/j.ophtha.2020.08.020","article-title":"Prevalence of keratoconus based on Scheimpflug imaging: the Raine study","volume":"128","author":"Chan","year":"2021","journal-title":"Opthalmology"},{"key":"ref3","doi-asserted-by":"publisher","first-page":"350","DOI":"10.3928\/1081-597X-20010501-09","article-title":"Complications of laser in situ keratomileusis: etiology, prevention, and treatment","volume":"17","author":"Ambrosio","year":"2001","journal-title":"J Refract Surg"},{"key":"ref4","doi-asserted-by":"publisher","first-page":"388","DOI":"10.1097\/01.ico.0000222479.68242.77","article-title":"Corneal ectasia after laser in situ keratomileusis in patients without apparent preoperative risk factors","volume":"25","author":"Klein","year":"2006","journal-title":"Cornea"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"440","DOI":"10.1016\/j.ophtha.2003.06.020","article-title":"Longitudinal study of the normal eyes in unilateral keratoconus patients","volume":"111","author":"Li","year":"2004","journal-title":"Ophthalmology"},{"key":"ref6","doi-asserted-by":"publisher","first-page":"585","DOI":"10.1016\/j.ajo.2009.10.021","article-title":"Long-term results of riboflavin ultraviolet a corneal collagen cross-linking for keratoconus in Italy: the Siena eye cross study","volume":"149","author":"Caporossi","year":"2010","journal-title":"Am J Ophthalmol"},{"key":"ref7","first-page":"2537","article-title":"Baseline findings in the collaborative longitudinal evaluation of keratoconus (CLEK) study","volume":"39","author":"Zadnik","year":"1998","journal-title":"Invest Ophthalmol Vis Sci"},{"key":"ref8","doi-asserted-by":"publisher","first-page":"359","DOI":"10.1097\/ICO.0000000000000408","article-title":"Group of Panelists for the global Delphi panel of keratoconus and Ectatic diseases. Global consensus on keratoconus and ectatic diseases","volume":"34","author":"Gomes","year":"2015","journal-title":"Cornea"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"794","DOI":"10.1097\/01.ico.0000226359.26678.d1","article-title":"A new method for grading the severity of keratoconus: the keratoconus severity score (KSS)","volume":"25","author":"McMahon","year":"2006","journal-title":"Cornea"},{"key":"ref10","doi-asserted-by":"publisher","first-page":"714","DOI":"10.1111\/ceo.14126","article-title":"Evaluation of artificial intelligence models for the detection of asymmetric keratoconus eyes using Scheimpflug tomography","volume":"50","author":"Zhe","year":"2022","journal-title":"Clin Experiment Ophthalmol"},{"key":"ref11","doi-asserted-by":"publisher","first-page":"1958","DOI":"10.1109\/TPAMI.2008.128","article-title":"80 million tiny images: a large data set for nonparametric object and scene recognition","volume":"30","author":"Torralba","year":"2008","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"ref12","first-page":"11","article-title":"Visualizing data using t-SNE","volume":"9","author":"Van der Maaten","year":"2008","journal-title":"J Mach Learn Res"},{"key":"ref13","volume-title":"TAO Toolkit","year":"2022"},{"key":"ref14","first-page":"6105","article-title":"Efficientnet: rethinking model scaling for convolutional neural networks","volume-title":"International Conference on Machine Learning","author":"Tan","year":"2019"},{"key":"ref15","doi-asserted-by":"publisher","first-page":"583","DOI":"10.1016\/j.patcog.2016.06.005","article-title":"Learning scale-variant and scale-invariant features for deep image classification","volume":"61","author":"Noord","year":"2017","journal-title":"Pattern Recogn"},{"key":"ref16","first-page":"1345","article-title":"A survey on transfer learning","volume":"10","author":"Pan","year":"2009","journal-title":"IEEE Trans Knowl Data Eng"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"3705","DOI":"10.2147\/OPTH.S425479","article-title":"Assessment of preoperative risk factors for post-LASIK ectasia development","volume":"17","author":"el-Naggar","year":"2023","journal-title":"Clin Ophthalmol"},{"key":"ref18","doi-asserted-by":"publisher","first-page":"364","DOI":"10.3928\/1081597X-20220601-01","article-title":"Evaluation of corneal biomechanical indices in distinguishing between Normal, very asymmetric, and bilateral Keratoconic eyes","volume":"38","author":"Herber","year":"2022","journal-title":"J Refract Surg"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"6609531","DOI":"10.1155\/2022\/6609531","article-title":"A hospital-based study on the prevalence of keratoconus in first-degree relatives of patients with keratoconus in Central China","volume":"2022","author":"Wang","year":"2022","journal-title":"J Ophthalmol"},{"key":"ref20","doi-asserted-by":"publisher","first-page":"580","DOI":"10.1097\/ICO.0000000000001537","article-title":"An early finding of keratoconus: increase in corneal densitometry","volume":"37","author":"Koc","year":"2018","journal-title":"Cornea"},{"key":"ref21","doi-asserted-by":"publisher","first-page":"259","DOI":"10.1016\/j.jcrs.2013.07.044","article-title":"Validation of metrics for the detection of PKC in a new patient collective","volume":"40","author":"B\u00fchren","year":"2014","journal-title":"J Cataract Refract Surg"},{"key":"ref22","doi-asserted-by":"publisher","first-page":"996","DOI":"10.1097\/ICO.0000000000000529","article-title":"Validation of an objective scoring system for Forme Fruste keratoconus detection and post-LASIK ectasia risk assessment in Asian eyes","volume":"34","author":"Chan","year":"2015","journal-title":"Cornea"},{"key":"ref23","doi-asserted-by":"publisher","first-page":"275","DOI":"10.1016\/j.jcrs.2015.09.020","article-title":"Accuracy of machine learning classifiers using bilateral data from a Scheimpflug camera for identifying eyes with preclinical signs of keratoconus","volume":"42","author":"Kov\u00e1cs","year":"2016","journal-title":"J Cataract Refract Surg"},{"key":"ref24","doi-asserted-by":"publisher","first-page":"510","DOI":"10.3928\/1081597X-20160523-01","article-title":"Combining Placido and Corneal Wavefront data for the detection of Forme Fruste keratoconus","volume":"32","author":"Saad","year":"2016","journal-title":"J Refract Surg"},{"key":"ref25","doi-asserted-by":"publisher","first-page":"827","DOI":"10.1097\/ICO.00000000000000834","article-title":"Evaluation of a machine-learning classifier for keratoconus detection based on Scheimpflug tomography","volume":"35","author":"Hidalgo","year":"2016","journal-title":"Cornea"},{"key":"ref26","doi-asserted-by":"publisher","first-page":"689","DOI":"10.1097\/ICO.00000000000001194","article-title":"Validation of an objective keratoconus detection system implemented in a Scheimpflug Tomographer and comparison with other methods","volume":"36","author":"Hidalgo","year":"2017","journal-title":"Cornea"},{"key":"ref27","doi-asserted-by":"publisher","first-page":"16486","DOI":"10.1038\/s41598-017-16568-y","article-title":"Characteristic of entire corneal topography and tomography for the detection of sub-clinical keratoconus with Zernike polynomials using Pentacam","volume":"7","author":"Zhe","year":"2017","journal-title":"Sci Rep"},{"key":"ref28","doi-asserted-by":"publisher","first-page":"434","DOI":"10.3928\/1081597X-20170426-02","article-title":"Integration of Scheimpflug-based corneal tomography and biomechanical assessments for enhancing ectasia detection","volume":"33","author":"Jr","year":"2017","journal-title":"J Refract Surg"},{"key":"ref29","doi-asserted-by":"publisher","first-page":"840","DOI":"10.3928\/1081597X-20181012-01","article-title":"Tomographic and biomechanical Scheimpflug imaging for keratoconus characterization: a validation of current indices","volume":"34","author":"Steinberg","year":"2018","journal-title":"J Refract Surg"},{"key":"ref30","doi-asserted-by":"publisher","first-page":"48","DOI":"10.1186\/s40662-020-00213-3","article-title":"Machine learning helps improve diagnostic ability of PKC using Scheimpflug and OCT imaging modalities","volume":"7","author":"Shi","year":"2020","journal-title":"Eye Vis"},{"key":"ref31","first-page":"5","article-title":"Keratoconus\/ectasia detection with the oculus pentacam: Belin\/Ambr\u00f3sio enhanced ectasia display","volume":"21","author":"Belin","year":"2007","journal-title":"Highlights Ophthalmol"},{"key":"ref32","doi-asserted-by":"publisher","first-page":"1158","DOI":"10.1097\/ICO.0000000000003038","article-title":"Kerato screen: early keratoconus classification with Zernike polynomial using deep learning","volume":"41","author":"Gao","year":"2022","journal-title":"Cornea"},{"key":"ref33","doi-asserted-by":"publisher","first-page":"53","DOI":"10.1167\/tvst.9.2.53","article-title":"Keratoconus screening based on deep learning approach of corneal topography","volume":"9","author":"Kuo","year":"2020","journal-title":"Transl Vis Sci Technol"}],"container-title":["Frontiers in Medicine"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2024.1458356\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,9,18]],"date-time":"2024-09-18T01:04:28Z","timestamp":1726621468000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2024.1458356\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,9,18]]},"references-count":33,"alternative-id":["10.3389\/fmed.2024.1458356"],"URL":"https:\/\/doi.org\/10.3389\/fmed.2024.1458356","relation":{},"ISSN":["2296-858X"],"issn-type":[{"value":"2296-858X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,9,18]]},"article-number":"1458356"}}