{"title":"Systematic review investigating the time taken for corneal stabilisation following contact lens cessation","authors":"Ali Abid, Himal Kandel, Stephanie Watson","doi":"10.1136/bjo-2025-327121","DOIUrl":"https://doi.org/10.1136/bjo-2025-327121","url":null,"abstract":"Background Contact lenses (CLs), used by approximately 140 million people globally, can induce corneal warpage, altering corneal geometry. This impacts corneal imaging, influencing diagnostic assessments, follow-ups and pre-operative evaluations. This systematic review aimed to determine the optimal CL cessation period before corneal imaging. Methods A systematic literature search was conducted across multiple databases. Studies were included if they assessed corneal stabilisation post-CL cessation by using corneal imaging techniques. Quality was rated using the modified Downs and Black checklist. Results Fifteen studies met the inclusion criteria; 13 prospective observational studies, one retrospective and one observational case series. Five of these studies included a control group. Following the removal of CLs, the included studies reported stabilisation of corneal parameters as follows: for soft CLs, 2–11.6 weeks (median 2.1 weeks); for hard CLs, 1–22 weeks (median 7.6 weeks); for orthokeratology lenses, 1–2 weeks (median 1.6 weeks) and for tinted lenses, 2–3 hours. Factors influencing stabilisation included lens material, water content, wear modality, CL fit and patient age. Two studies were found to be of excellent methodological quality, 11 were assessed as good and two as fair. No studies were found to be of poor quality. Conclusion Corneal stabilisation time post-CL cessation was variable and dependent on the lens type and fit, patient age and method of determining corneal stabilisation. Personalised recommendations and standardised assessments are essential for optimising diagnostic and surgical outcomes. Further research is needed to develop comprehensive guidelines for consistent CL removal protocols. PROSPERO registration number CRD42023484401. Data sharing not applicable as no datasets generated and/or analysed for this study.","PeriodicalId":9313,"journal":{"name":"British Journal of Ophthalmology","volume":"16 1","pages":"bjo-2025-327121"},"PeriodicalIF":4.1,"publicationDate":"2025-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144145998","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Retrospective multicentre study on functional vision score calculation using automated perimeter","authors":"Mieko Tsuruoka, Yoshimune Hiratsuka, Kenji Inoue, Miki Murakami, Kazuma Oku, Hiroyuki Kondo, Mayumi Sainohira, Taiji Sakamoto, Akira Murakami","doi":"10.1136/bjo-2025-327122","DOIUrl":"https://doi.org/10.1136/bjo-2025-327122","url":null,"abstract":"Background/aims This study aimed to develop a predictive equation for Functional Field Score (FFS), a component of the Functional Vision Score (FVS) derived from binocular Esterman test results via automated perimetry (AP) and to evaluate agreement with Goldmann perimetry (GP)-based FFS. Methods This retrospective multicentre study included 240 patients (mean age: 61±17.5 years) who underwent both GP and AP testing at four Japanese institutions from January 2019 to October 2023. Binocular Esterman visual field testing was conducted using the Humphrey Field Analyser. After conducting a correlation analysis using Pearson’s correlation coefficient, we derived a predictive equation for GP-based FFS from AP-based FFS using linear regression analysis. Bland-Altman analysis assessed agreement between predicted AP-FFS and GP-FFS. Results A strong correlation was observed between FFS based on GP and the number of points seen in the binocular Esterman test (r=0.77; 95% CI 0.71 to 0.81). The predictive equation for FFS was: FFS=(the number of points seen)×0.41+37.0. Agreement between predicted and GP-based FFS was high, with a mean difference of 0.12 (95% CI −21.63 to 21.88). Conclusions AP-based FFS provides a reliable alternative to GP-based FVS, enabling effective assessment of visual impairment in environments lacking GP accessibility. This approach may enhance the practical evaluation of functional vision, particularly in clinical or research settings where AP is more widely available. No data are available.","PeriodicalId":9313,"journal":{"name":"British Journal of Ophthalmology","volume":"46 1","pages":""},"PeriodicalIF":4.1,"publicationDate":"2025-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144133408","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jurgen Sota, Luciana Breda, Maria Pia Paroli, Soad Hashad, Alejandra de-la-Torre, Silvana Guerriero, Abdurrahman Tufan, Juanita Cardona-López, María José Altamar-García, Saverio La Bella, Antonio Vitale, Edoardo Biancalana, Emanuela Del Giudice, Francesca Minoia, Sulaiman Al-Mayouf, Bruno Frediani, Luca Cantarini, Claudia Fabiani
{"title":"Tocilizumab effectiveness in paediatric non-infectious uveitis: data from the International AIDA Network Registries on ocular inflammatory disorders","authors":"Jurgen Sota, Luciana Breda, Maria Pia Paroli, Soad Hashad, Alejandra de-la-Torre, Silvana Guerriero, Abdurrahman Tufan, Juanita Cardona-López, María José Altamar-García, Saverio La Bella, Antonio Vitale, Edoardo Biancalana, Emanuela Del Giudice, Francesca Minoia, Sulaiman Al-Mayouf, Bruno Frediani, Luca Cantarini, Claudia Fabiani","doi":"10.1136/bjo-2025-327410","DOIUrl":"https://doi.org/10.1136/bjo-2025-327410","url":null,"abstract":"To describe tocilizumab (TCZ) effectiveness in 15 children with refractory non-infectious uveitis. Reported outcomes are the number of relapses before and after treatment, steroid-sparing effect and drug retention rate. Macular oedema, fluorangiographic findings and ocular complications are also reported. The mean number of ocular relapses significantly decreased from 314 per 100 eyes/year to 106 per 100 eyes/year (p=0.016). A significant steroid-sparing effect was detected (p=0.037). TCZ drug survival was 77.4% at 6 months, followed by 61.9% at 12, 24 and 36 months of follow-up. Macular oedema and retinal vasculitis resolved in all affected eyes. Data are available upon reasonable request. The data underlying this article will be shared upon reasonable request to the corresponding author.","PeriodicalId":9313,"journal":{"name":"British Journal of Ophthalmology","volume":"34 1","pages":""},"PeriodicalIF":4.1,"publicationDate":"2025-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144133407","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Enhancing pathological myopia diagnosis: a bimodal artificial intelligence approach integrating fundus and optical coherence tomography imaging for precise atrophy, traction and neovascularisation grading.","authors":"Zhiyan Xu,Yajie Yang,Huan Chen,Ruo'an Han,Xiaoxu Han,Jianchun Zhao,Weihong Yu,Zhikun Yang,Youxin Chen","doi":"10.1136/bjo-2024-326252","DOIUrl":"https://doi.org/10.1136/bjo-2024-326252","url":null,"abstract":"BACKGROUNDPathological myopia (PM) has emerged as a leading cause of global visual impairment, early detection and precise grading of PM are crucial for timely intervention. The atrophy, traction and neovascularisation (ATN) system is applied to define PM progression and stages with precision. This study focuses on constructing a comprehensive PM image dataset comprising both fundus and optical coherence tomography (OCT) images and developing a bimodal artificial intelligence (AI) classification model for ATN grading in PM.METHODSThis single-centre retrospective cross-sectional study collected 2760 colour fundus photographs and matching OCT images of PM from January 2019 to November 2022 at Peking Union Medical College Hospital. Ophthalmology specialists labelled and inspected all paired images using the ATN grading system. The AI model used a ResNet-50 backbone and a multimodal multi-instance learning module to enhance interaction across instances from both modalities.RESULTSPerformance comparisons among single-modality fundus, OCT and bimodal AI models were conducted for ATN grading in PM. The bimodality model, dual-deep learning (DL), demonstrated superior accuracy in both detailed multiclassification and biclassification of PM, which aligns well with our observation from instance attention-weight activation maps. The area under the curve for severe PM using dual-DL was 0.9635 (95% CI 0.9380 to 0.9890), compared with 0.9359 (95% CI 0.9027 to 0.9691) for the solely OCT model and 0.9268 (95% CI 0.8915 to 0.9621) for the fundus model.CONCLUSIONSOur novel bimodal AI multiclassification model for PM ATN staging proves accurate and beneficial for public health screening and prompt referral of PM patients.","PeriodicalId":9313,"journal":{"name":"British Journal of Ophthalmology","volume":"28 1","pages":""},"PeriodicalIF":4.1,"publicationDate":"2025-05-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144103737","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jorge Ruiz-Medrano, Patricia Udaondo Mirete, Marina Fernández-Jiménez, Monica Asencio-Duran, José Ignacio Fernández-Vigo, Marta Medina-Baena, Ignacio Flores-Moreno, Jesus Pareja-Esteban, Sara Touhami, Audrey Giocanti-Aurégan, Maria Vittoria Cicinelli, Anat Loewenstein, José M Ruiz-Moreno
{"title":"Biomarkers of risk of switching to dexamethasone implant for the treatment of diabetic macular oedema in real clinical practice: a multicentric study","authors":"Jorge Ruiz-Medrano, Patricia Udaondo Mirete, Marina Fernández-Jiménez, Monica Asencio-Duran, José Ignacio Fernández-Vigo, Marta Medina-Baena, Ignacio Flores-Moreno, Jesus Pareja-Esteban, Sara Touhami, Audrey Giocanti-Aurégan, Maria Vittoria Cicinelli, Anat Loewenstein, José M Ruiz-Moreno","doi":"10.1136/bjo-2024-325665","DOIUrl":"https://doi.org/10.1136/bjo-2024-325665","url":null,"abstract":"Objective To establish the influence of different optical coherence tomography (OCT) biomarkers at baseline treatment on the potential response to anti-vascular endothelial growth factor (VEGF) treatment for diabetic macular oedema (DME). Methods Multicentric, retrospective, case-series study in patients with DME switched to dexamethasone implant injections (DEX-i) after anti-VEGF in real clinical practice. Biomarkers analysed on OCT images at baseline included intraretinal fluid (IRF), subretinal fluid (SRF), disorganisation of retinal inner layers (DRIL), disorganisation of retinal outer layers (DROL), hyperreflective foci (HRF), hyperreflective cystoid walls (HCW), dense intraretinal cyst (DIR) and vitreomacular interface (VMI) abnormalities. DME was classified according to the European School for Advanced Studies in Ophthalmology classification. Patients who were treated with anti-VEGF injections with an adequate response were selected as the control group. Results 275 eyes were analysed in this study; 209 eyes (76.0%) switched from anti-VEGF to DEX-i were compared with 66 control eyes (24.0%). Patients who required switching were statistically older, showed worse initial BCVA and higher CRT. Logistic regression analyses showed that female gender, age, central retinal thickness, type of diabetes, SRF, HCW, DIR and VMI increase the likelihood of switching. The OR regarding the need to switch generated by the presence of two of these three factors (SRF, HCW, VMI) was 48.95. Having all three multiplies it by 4.56×1016. Conclusion If baseline OCT shows two of SRF, HCW and VMI biomarkers at baseline, the risk of failure of anti-VEGF therapy is close to 50%. In the presence at baseline of all three biomarkers, failure of anti-VEGF therapy is almost certain. Data are available upon reasonable request.","PeriodicalId":9313,"journal":{"name":"British Journal of Ophthalmology","volume":"9 1","pages":""},"PeriodicalIF":4.1,"publicationDate":"2025-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144067131","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Nehal Nailesh Mehta, Ines Doris Nagel, Akshay Agnihotri, Fritz Gerald Paguiligan Kalaw, Anna Heinke, Lingyun Cheng, Dirk-Uwe Bartsch, William R Freeman
{"title":"Anti-VEGF injections and macular atrophy progression in patients with neovascular age-related macular degeneration in remission","authors":"Nehal Nailesh Mehta, Ines Doris Nagel, Akshay Agnihotri, Fritz Gerald Paguiligan Kalaw, Anna Heinke, Lingyun Cheng, Dirk-Uwe Bartsch, William R Freeman","doi":"10.1136/bjo-2024-326124","DOIUrl":"https://doi.org/10.1136/bjo-2024-326124","url":null,"abstract":"Aims To determine the effect of continuing anti-vascular endothelial growth factor (VEGF) injections on the progression of macular atrophy (MA) during remission of neovascular age-related macular degeneration (nAMD). Methods In this retrospective cohort study, 59 eyes with nAMD with at least 6-month remission (disease inactivity) were analysed and were grouped into two. In group 1, anti-VEGF injections were stopped after remission (holiday). In group 2, injections were continued despite inactivity (maintenance). Using blue autofluorescence images via Heidelberg Spectralis, MA area was measured at initial injection, remission onset and the latest available remission point. The absence of subretinal haemorrhage, intraretinal fluid, subretinal fluid or subretinal hyper-reflective material associated with fluid on serial spectral domain optical coherence tomography scans was used to confirm the inactivity of the disease (remission). The rate of progression of MA during the period of remission was measured for the two groups. Results In group 1, 30 eyes received a mean of 16.97 injections over 39.2 months, followed by 21 months of drug holiday. In group 2, 29 eyes received a mean of 27.1 injections over 62.16 months, followed by a mean of 11.59 injections over 19.32 months for maintenance. The MA in the maintenance group progressed faster than the holiday group during remission (p=0.03). Conclusions Maintenance injections for nAMD in remission significantly increase progression of MA. Data are available on reasonable request. Data are available on reasonable request to WRF.","PeriodicalId":9313,"journal":{"name":"British Journal of Ophthalmology","volume":"35 1","pages":""},"PeriodicalIF":4.1,"publicationDate":"2025-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144066816","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jae-Seung Yun, Sang-Hyuk Jung, Seung Min Jung, Hong-Hee Won, Dokyoon Kim, Jin A Choi
{"title":"Corneal hysteresis as a biomarker in glaucoma development among patients with myopia: a prospective study based on the UK Biobank","authors":"Jae-Seung Yun, Sang-Hyuk Jung, Seung Min Jung, Hong-Hee Won, Dokyoon Kim, Jin A Choi","doi":"10.1136/bjo-2024-326817","DOIUrl":"https://doi.org/10.1136/bjo-2024-326817","url":null,"abstract":"Background/aims We investigated the impact of potential risk factors on the development of incident glaucoma in individuals with myopia. Methods This prospective cohort study investigated participants aged 40–69 years from the UK Biobank. Participants were classified based on the degree of myopia: non-myopic (spherical equivalent (SE)≤−0.5D), mild myopia (−3.0D<SE≤−0.5 D), moderate myopia (−6.0 D<SE≤−3.0 D) and high myopia (SE≤−6.0D). Cox proportional-hazard models assessed the relationship of ocular and systemic risk factors with new glaucoma events across the myopia subgroups. Results Of 105 548 participants who did not have prior glaucoma, 35 870 were classified as myopic and 69 678 as non-myopic. A total of 1877 incident glaucoma cases (1.78%) were documented during a median 11.1 years of follow-up. Those who developed glaucoma had lower SE, lower corneal hysteresis (CH) and higher intraocular pressure. Systemic factors such as type 2 diabetes, dyslipidaemia, baseline fasting blood sugar and haemoglobin A1c were significantly higher in those who developed glaucoma, but only in the non-myopia and mild myopia subgroups. Degree of myopia and low CH (≤10.1 mm Hg) were significantly associated with increased glaucoma incidence. When we explored the impact of CH on the risk of incident glaucoma according to myopia grade, consistent elevation of risk with low CH was observed, with the highest risk being exhibited in the high myopia population (HR 1.67 (95% CI 1.10 to 2.57)). Conclusions Our results demonstrated a significant joint association between CH and myopia regarding glaucoma development, suggesting that altered ocular biomechanical properties could play a pronounced role in glaucoma development, particularly within the highly myopic population. Data may be obtained from a third party and are not publicly available.","PeriodicalId":9313,"journal":{"name":"British Journal of Ophthalmology","volume":"196 1","pages":""},"PeriodicalIF":4.1,"publicationDate":"2025-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143979441","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Katherine Overbey,Francesco Romano,Xinyi Ding,Cade F Bennett,Isabella Stettler,Itika Garg,Ioanna Ploumi,Filippos Vingopoulos,Melissa Yuan,Peyman Razavi,Matthew Finn,Inês Laíns,Nimesh A Patel,Leo Kim,David Wu,Dean Eliott,Deeba Husain,Demetrios Vavvas,Joan W Miller,John B Miller
{"title":"Choriocapillaris impairment in dry AMD: insights from swept-source OCT angiography and associations with structural biomarkers.","authors":"Katherine Overbey,Francesco Romano,Xinyi Ding,Cade F Bennett,Isabella Stettler,Itika Garg,Ioanna Ploumi,Filippos Vingopoulos,Melissa Yuan,Peyman Razavi,Matthew Finn,Inês Laíns,Nimesh A Patel,Leo Kim,David Wu,Dean Eliott,Deeba Husain,Demetrios Vavvas,Joan W Miller,John B Miller","doi":"10.1136/bjo-2024-326416","DOIUrl":"https://doi.org/10.1136/bjo-2024-326416","url":null,"abstract":"AIMSTo assess choriocapillaris flow deficit percentage (CCFD%) across stages of dry age-related macular degeneration (AMD) using swept-source optical coherence tomography angiography (SS-OCTA).METHODSThis cross-sectional, observational study included 270 eyes (182 patients), classified as early (70 eyes), intermediate (121 eyes) and geographic atrophy (GA, 79 eyes).Participants underwent a complete examination including macular 6×6 mm SS-OCTA scans (PLEX Elite 9000). Scans were reviewed and analysed for subretinal drusenoid deposits (SDDs), retinal pigment epithelium (RPE) atrophy size, incomplete RPE and outer retinal atrophy (iRORA) and drusen volume (3 mm). CCFD% was calculated after compensation and binarisation using Phansalkar's method (r=4-15 pixels) in various early treatment for diabetic retinopathy study sectors. Linear mixed-effects models adjusted for age evaluated associations with AMD stages and other imaging biomarkers.RESULTSCCFD% progressively increased with advancing dry AMD stages. Intermediate AMD eyes showed higher CCFD% than early AMD ones across all regions (p<0.001). GA eyes exhibited significantly higher CCFD% compared with early (p<0.001) and intermediate AMD eyes (p<0.001).SDDs were significantly associated with higher CCFD% in early (p<0.01) and intermediate AMD (p<0.05) for almost all regions examined, but not in GA (p>0.05). iRORA presence in iAMD and larger RPE atrophy in GA correlated with increased CCFD% (p<0.001).CONCLUSIONSThis study provides a comprehensive reference database for CCFD% across the stages of dry AMD using SS-OCTA. CCFD% increased with AMD severity, iRORA, SDDs, particularly in early and intermediate stages, and RPE atrophy size. Our findings support CCFD% as a valuable biomarker for clinical and research applications, warranting longitudinal studies to validate its prognostic value.","PeriodicalId":9313,"journal":{"name":"British Journal of Ophthalmology","volume":"12 1","pages":""},"PeriodicalIF":4.1,"publicationDate":"2025-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143945239","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Validation of a self-administered Home ETDRS visual acuity testing (H-ETDRS) for self-monitoring vision changes in retinal diseases.","authors":"Zitian Liu,Yanjing Huang,Ling Jin,Caineng Pan,Xiaotong Han,Yingfeng Zheng","doi":"10.1136/bjo-2024-326283","DOIUrl":"https://doi.org/10.1136/bjo-2024-326283","url":null,"abstract":"OBJECTIVETo validate a self-administered Home Early Treatment Diabetic Retinopathy Study (H-ETDRS) visual acuity (VA) test and to assess its accuracy in detecting VA decline in patients with retinal diseases.METHODSA validation group of 156 participants and a testing group of 100 participants with diabetic retinopathy and age-related macular degeneration were recruited. All participants underwent monocular distance best-corrected VA (BCVA) tests in the study eye using the standard ETDRS (S-ETDRS) charts in clinic and the H-ETDRS device in a simulated home setting. Participants in the validation group repeated both methods. H-ETDRS letter scores were compared with S-ETDRS scores, evaluating mean difference and test-retest reliability (TRR). Intraclass correlation coefficients (ICCs) were calculated to measure the agreement. Participants in the testing group repeated both methods with a Bangerter filter density of 0.6 to simulate BCVA change. The performance of H-ETDRS in detecting acuity change was evaluated, with S-ETDRS serving as the standard.RESULTSFor the validation group, the mean difference between S-ETDRS and H-ETDRS scores was -1.38 letters (95% limits of agreement (LOA) -6.99 to 4.22). The scores of the two tests were highly correlated with an ICC of 0.98. The mean differences between initial test and retest scores were 0.11 letters (95% LOA -2.62 to 2.84) for S-ETDRS and -0.13 letters (95% LOA -5.64 to 5.38) for H-ETDRS. The ICCs of test-retest letter scores were 1.00 for S-ETDRS and 0.98 for H-ETDRS. For participants in the testing group who showed BCVA decline of one line or more with Bangerter filter, 96.7% were able to detect VA changes using H-ETDRS.CONCLUSIONS AND RELEVANCEThe H-ETDRS test demonstrates high TRR and good concordance with S-ETDRS test for retinal patients. The H-ETDRS is viable for self-monitoring VA changes in retinal diseases, however, further validation studies in actual home settings are warranted.","PeriodicalId":9313,"journal":{"name":"British Journal of Ophthalmology","volume":"26 1","pages":""},"PeriodicalIF":4.1,"publicationDate":"2025-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143932828","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Wanlin Fan,Martine Johanna Jager,Weiwei Dai,Ludwig M Heindl
{"title":"Deep learning-based system for automatic identification of benign and malignant eyelid tumours.","authors":"Wanlin Fan,Martine Johanna Jager,Weiwei Dai,Ludwig M Heindl","doi":"10.1136/bjo-2025-327127","DOIUrl":"https://doi.org/10.1136/bjo-2025-327127","url":null,"abstract":"AIMSOur aim is to develop a deep learning-based system for automatically identifying and classifying benign and malignant tumours of the eyelid to improve diagnostic accuracy and efficiency.METHODSThe dataset includes photographs of normal eyelids, benign and malignant eyelid tumours and was randomly divided into a training and validation dataset in a ratio of 8:2. We used the training dataset to train eight convolutional neural network models to classify normal eyelids, benign and malignant eyelid tumours. These models included VGG16, ResNet50, Inception-v4, EfficientNet-V2-M and their variants. The validation dataset was used to evaluate and compare the performance of the different deep learning models.RESULTSAll eight models achieved an average accuracy greater than 0.746 for identifying normal eyelids, benign and malignant eyelid tumours, with an average sensitivity and specificity exceeding 0.790 and 0.866, respectively. The mean area under the receiver operating characteristic curve (AUC) for the eight models was more than 0.904 in correctly identifying normal eyelids, benign and malignant eyelid tumours. The dual-path Inception-v4 network demonstrated the highest performance, with an AUC of 0.930 (95% CI 0.900 to 0.954) and an F1-score of 0.838 (95% CI 0.787 to 0.882).CONCLUSIONThe deep learning-based system shows significant potential in improving the diagnosis of eyelid tumours, providing a reliable and efficient tool for clinical practice. Future work will validate the model with more extensive and diverse datasets and integrate it into clinical workflows for real-time diagnostic support.","PeriodicalId":9313,"journal":{"name":"British Journal of Ophthalmology","volume":"35 1","pages":""},"PeriodicalIF":4.1,"publicationDate":"2025-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143932817","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}