Dataset Information

Evaluating the impact of vitreomacular adhesion on anti-VEGF therapy for retinal vein occlusion using machine learning.

ABSTRACT: Vitreomacular adhesion (VMA) represents a prognostic biomarker in the management of exudative macular disease using anti-vascular endothelial growth factor (VEGF) agents. However, manual evaluation of VMA in 3D optical coherence tomography (OCT) is laborious and data on its impact on therapy of retinal vein occlusion (RVO) are limited. The aim of this study was to (1) develop a fully automated segmentation algorithm for the posterior vitreous boundary and (2) to study the effect of VMA on anti-VEGF therapy for RVO. A combined machine learning/graph cut segmentation algorithm for the posterior vitreous boundary was designed and evaluated. 391 patients with central/branch RVO under standardized ranibizumab treatment for 6/12 months were included in a systematic post-hoc analysis. VMA (70%) was automatically differentiated from non-VMA (30%) using the developed method combined with unsupervised clustering. In this proof-of-principle study, eyes with VMA showed larger BCVA gains than non-VMA eyes (BRVO: 15?±?12 vs. 11?±?11 letters, p?=?0.02; CRVO: 18?±?14 vs. 9?±?13 letters, p?

SUBMITTER: Waldstein SM

PROVIDER: S-EPMC5462785 | biostudies-literature | 2017 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Evaluating the impact of vitreomacular adhesion on anti-VEGF therapy for retinal vein occlusion using machine learning.

Waldstein Sebastian M SM Montuoro Alessio A Podkowinski Dominika D Philip Ana-Maria AM Gerendas Bianca S BS Bogunovic Hrvoje H Schmidt-Erfurth Ursula U

Scientific reports 20170607 1

Vitreomacular adhesion (VMA) represents a prognostic biomarker in the management of exudative macular disease using anti-vascular endothelial growth factor (VEGF) agents. However, manual evaluation of VMA in 3D optical coherence tomography (OCT) is laborious and data on its impact on therapy of retinal vein occlusion (RVO) are limited. The aim of this study was to (1) develop a fully automated segmentation algorithm for the posterior vitreous boundary and (2) to study the effect of VMA on anti-V ...[more]

PMID: 28592811

Similar Datasets

Project description:BackgroundRetinal vein occlusion (RVO) is one of the most frequent retinal vascular diseases. In this study, we aimed to investigate the predictive factors of visual outcome for RVO patients who underwent anti-vascular endothelial growth factor (VEGF) therapy.MethodsRVO patients who underwent anti-VEGF treatment were recruited in this study from January 2018 to June 2020. Clinical data and optical coherence tomography (OCT) parameters were retrospectively reviewed. Best-corrected visual acuity (BCVA) was examined at baseline and after anti-VEGF therapy. Predictive factors associated with visual outcome were assessed by logistic regression model. Treatment-related adverse events were also recorded.ResultsThe average logMAR BCVA was 0.91 at baseline and 0.70 at final examination (P = 0.003). Among 75 patients, 41 experienced visual improvement were categorized as group A, the remaining 34 patients without improved vision were categorized as group B. Patients in group A demonstrated better visual outcomes, including decreased logMAR BCVA (average logMAR BCVA: 0.53 in group A vs. 0.91 in group B, P < 0.001) and central retinal thickness (CRT) (average CRT: 230.88 µm in group A vs. 404.97 µm in group B, P < 0.001) after anti-VEGF treatment. Multivariable analysis showed that injection frequency (odds ratio [OR], 2.623; 95% confidence interval [CI], [1.282-5.366]), hypertension (odds ratio [OR], 0.189; 95% CI [0.044-0.811]), hyperlipemia (odds ratio [OR], 0.195; 95% CI [0.040-0.941]) and external limiting membrane (ELM) disruption (odds ratio [OR], 0.148; 95% CI [0.032-0.691]) were all significantly associated with the visual outcome of RVO patients who underwent anti-VEGF treatment. In general, anti-VEGF therapy was feasible for all RVO patients, though the response to anti-VEGF was suboptimal in certain patients. Prognostic factors including injection frequency, hypertension, hyperlipemia and ELM disruption may all be useful to provide predictive information of visual outcome of RVO patients in response to anti-VEGF treatment.

Project description:BackgroundSD-OCT is becoming commonplace in everyday practice. Vitreomacular adhesions (VMAs) are being more routinely diagnosed. Predictive studies to the natural course of VMA are thus clinically significant. Spectral domain-optical coherence tomography (SD-OCT) was presently utilized to analyze the incidence of floaters, the complete vitreomacular separation or VMA, the VMA complication, the vitreomacular angle (VMAng), and the complication mechanism.MethodsMonthly SD-OCT was performed on patients with/without symptomatic floaters. OCT allowed VMA and vitreomacular separation to be compared. The incidence was assessed applying one-tailed Fisher's exact tests. The VMAngs between the inner retina and posterior hyaloid were measured, and the complication mechanism was studied using OCT image. For macular hole (MH), pre- and/or post-operative best corrected visual acuities (BCVAs; LogMAR), refractions and photoreceptor conditions were also evaluated.ResultsTotally, 124 eyes were included; there were 116 eyes with VMA and 8 eyes with vitreomacular separation. Considering the percentages over 124 eyes, floaters were present in 14.5% of enrolled eyes (=18/124), consisting of 12.9% of eyes with VMA (16/124) and 1.6% of eyes with vitreomacular separation (2/124). Moreover, there were twelve eyes (9.7%) with VMA-associated vision-threatening complications, including MH (n = 8; 6.5%), retinal detachment (RD; n = 2; 1.6%), vitreomacular traction (VMT; n = 1; 0.8%) and macular pucker (MP; n = 1; 0.8%). Eyes with initial VMA had a significantly greater possibility of complications than eyes with initial vitreomacular separation (p = 0.03). Among these eyes with MH (n = 8), the pre-operative BCVA (LogMAR) was 1.1 ± 0.5, which was insignificantly (p = 0.35) improved to 0.8 ± 0.7 post-operatively. The VMAng of VMA eyes with MHs was 24.2 ± 24.9° (n = 8). The critical VMAng was 13.3°.ConclusionsA minority of eyes with VMA or vitreomacular separation had floaters. Moreover, the use of SD-OCT could identify vision-threatening sequelae, namely MH, RD, MP and VMT, and this was significantly more frequent in eyes with VMA than in eyes with complete vitreomacular separation. Therefore, SD-OCT might be a useful way of identifying either identity, and evaluating VMA-associated complications. Whether VMA eyes with MH (n = 8) that have a VMAng greater than critical VMAng have a greater likelihood of tangential traction and subsequent MH needs further investigation.

Dataset Information

Evaluating the impact of vitreomacular adhesion on anti-VEGF therapy for retinal vein occlusion using machine learning.

Publications

Evaluating the impact of vitreomacular adhesion on anti-VEGF therapy for retinal vein occlusion using machine learning.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets