The Scheimpflug-derived shape and biomechanical parameters are able to accurately distinguish normal corneas from frank (clinical) keratoconic corneas. However, the combined parameters were more effective. Further studies should test milder ectasia cases.
ObjectiveTo evaluate the accuracy of convolutional neural networks technique (CNN) in detecting keratoconus using colour-coded corneal maps obtained by a Scheimpflug camera.DesignMulticentre retrospective study.Methods and analysisWe included the images of keratoconic and healthy volunteers’ eyes provided by three centres: Royal Liverpool University Hospital (Liverpool, UK), Sedaghat Eye Clinic (Mashhad, Iran) and The New Zealand National Eye Center (New Zealand). Corneal tomography scans were used to train and test CNN models, which included healthy controls. Keratoconic scans were classified according to the Amsler-Krumeich classification. Keratoconic scans from Iran were used as an independent testing set. Four maps were considered for each scan: axial map, anterior and posterior elevation map, and pachymetry map.ResultsA CNN model detected keratoconus versus health eyes with an accuracy of 0.9785 on the testing set, considering all four maps concatenated. Considering each map independently, the accuracy was 0.9283 for axial map, 0.9642 for thickness map, 0.9642 for the front elevation map and 0.9749 for the back elevation map. The accuracy of models in recognising between healthy controls and stage 1 was 0.90, between stages 1 and 2 was 0.9032, and between stages 2 and 3 was 0.8537 using the concatenated map.ConclusionCNN provides excellent detection performance for keratoconus and accurately grades different severities of disease using the colour-coded maps obtained by the Scheimpflug camera. CNN has the potential to be further developed, validated and adopted for screening and management of keratoconus.
Purpose: To investigate corneal biomechanical response parameters in varying degrees of myopia and their correlation with corneal geometrical parameters and axial length.Methods: In this prospective cross-sectional study, 172 eyes of 172 subjects, the severity degree of myopia was categorized into mild, moderate, severe, and extreme myopia. Cycloplegic refraction, corneal tomography using Pentacam HR, corneal biomechanical assessment using Corvis ST and Ocular Response Analyser (ORA), and ocular biometry using IOLMaster 700 were performed for all subjects. A general linear model was used to compare biomechanical parameters in various degrees of myopia, while central corneal thickness (CCT) and biomechanically corrected intraocular pressure (bIOP) were considered as covariates. Multiple linear regression was used to investigate the relationship between corneal biomechanical parameters with spherical equivalent (SE), axial length (AXL), bIOP, mean keratometry (Mean KR), and CCT.Results: Corneal biomechanical parameters assessed by Corvis ST that showed significant differences among the groups were second applanation length (AL2, p = 0.035), highest concavity radius (HCR, p < 0.001), deformation amplitude (DA, p < 0.001), peak distance (PD, p = 0.022), integrated inverse radius (IR, p < 0.001) and DA ratio (DAR, p = 0.004), while there were no significant differences in the means of pressure-derived parameters of ORA between groups. Multiple regression analysis showed all parameters of Corvis ST have significant relationships with level of myopia (SE, AXL, Mean KR), except AL1 and AL2. Significant biomechanical parameters showed progressive reduction in corneal stiffness with increasing myopia (either with greater negative SE or greater AXL), independent of IOP and CCT. Also, corneal hysteresis (CH) or ability to dissipate energy from the ORA decreased with increasing level of myopia.Conclusions: Dynamic corneal response assessed by Corvis ST shows evidence of biomechanical changes consistent with decreasing stiffness with increasing levels of myopia in multiple parameters. The strongest correlations were with highest concavity parameters where the sclera influence is maximal.
Purpose:To evaluate changes in corneal topography and biomechanical properties after collagen cross-linking (CXL) for progressive keratoconus.Patients and Methods:Collagen cross-linking was performed on 97 eyes. We assessed uncorrected visual acuity (UCVA) and best corrected visual acuity (BCVA). Corneal topography indices were evaluated using placido disc topography, scanning slit anterior topography (Orbscan II), and rotating Scheimpflug topography (Pentacam). Specular microscopy and corneal biomechanics were evaluated.Results:A 1-year-follow-up results revealed that UCVA improved from 0.31 to 0.45 and BCVA changed from 0.78 to 0.84 (P < 0.001). The mean of average keratometry value decreased from 49.62 to 47.95 D (P < 0.001). Astigmatism decreased from 4.84 to 4.24 D (P < 0.001). Apex corneal thickness decreased from 458.11 to 444.46 μm. Corneal volume decreased from 56.66 to 55.97 mm3 (P < 0.001). Posterior best fit sphere increased from 55.50 to 46.03 mm (P = 0.025). Posterior elevation increased from 99.2 to 112.22 μm (P < 0.001). Average progressive index increased from 2.26 to 2.56 (P < 0.001). A nonsignificant decrease was observed in mean endothelial count from 2996 to 2928 cell/mm2 (P = 0.190). Endothelial coefficient of variation (CV) increased nonsignificantly from 18.26 to 20.29 (P = 0.112). Corneal hysteresis changed from 8.18 to 8.36 (P = 0.552) and corneal resistance factor increased from 6.98 to 7.21 (P = 0.202), so these changes were not significant.Conclusion:Visual acuity and K values improved after CXL. In spite of the nonsignificant increase in endothelial cell count and increase in the CV, CLX seems to be a safe treatment for keratoconus. Further studies with larger sample sizes and longer follow-up periods are recommended.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.