1
|
Hartmann LM, Langhans DS, Eggarter V, Freisenich TJ, Hillenmayer A, König SF, Vounotrypidis E, Wolf A, Wertheimer CM. Keratoconus Progression Determined at the First Visit: A Deep Learning Approach With Fusion of Imaging and Numerical Clinical Data. Transl Vis Sci Technol 2024; 13:7. [PMID: 38727695 PMCID: PMC11104256 DOI: 10.1167/tvst.13.5.7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/21/2023] [Accepted: 03/15/2024] [Indexed: 05/22/2024] Open
Abstract
Purpose Multiple clinical visits are necessary to determine progression of keratoconus before offering corneal cross-linking. The purpose of this study was to develop a neural network that can potentially predict progression during the initial visit using tomography images and other clinical risk factors. Methods The neural network's development depended on data from 570 keratoconus eyes. During the initial visit, numerical risk factors and posterior elevation maps from Scheimpflug imaging were collected. Increase of steepest keratometry of 1 diopter during follow-up was used as the progression criterion. The data were partitioned into training, validation, and test sets. The first two were used for training, and the latter for performance statistics. The impact of individual risk factors and images was assessed using ablation studies and class activation maps. Results The most accurate prediction of progression during the initial visit was obtained by using a combination of MobileNet and a multilayer perceptron with an accuracy of 0.83. Using numerical risk factors alone resulted in an accuracy of 0.82. The use of only images had an accuracy of 0.77. The most influential risk factors in the ablation study were age and posterior elevation. The greatest activation in the class activation maps was seen at the highest posterior elevation where there was significant deviation from the best fit sphere. Conclusions The neural network has exhibited good performance in predicting potential future progression during the initial visit. Translational Relevance The developed neural network could be of clinical significance for keratoconus patients by identifying individuals at risk of progression.
Collapse
Affiliation(s)
| | | | | | | | - Anna Hillenmayer
- Department of Ophthalmology, University Hospital Ulm, Ulm, Germany
| | - Susanna F. König
- Department of Ophthalmology, University Hospital Ulm, Ulm, Germany
| | | | - Armin Wolf
- Department of Ophthalmology, University Hospital Ulm, Ulm, Germany
| | | |
Collapse
|
2
|
Zheng X, Xin Y, Wang C, Fan Y, Yang P, Li L, Yin D, Zhang E, Hong Y, Bao H, Wang J, Bao F, Zhang W, Chen S, Elsheikh A, Swain M. Use of Nanoindentation in Determination of Regional Biomechanical Properties of Rabbit Cornea After UVA Cross-Linking. Invest Ophthalmol Vis Sci 2023; 64:26. [PMID: 37850947 PMCID: PMC10593136 DOI: 10.1167/iovs.64.13.26] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/06/2023] [Accepted: 09/25/2023] [Indexed: 10/19/2023] Open
Abstract
Purpose To evaluate the regional effects of different corneal cross-linking (CXL) protocols on corneal biomechanical properties. Methods The study involved both eyes of 50 rabbits, and the left eyes were randomized to the five intervention groups, which included the standard CXL group (SCXL), which was exposed to 3-mW/cm2 irradiation, and three accelerated CXL groups (ACXL1-3), which were exposed to ultraviolet-A at irradiations of 9 mW/cm2, 18 mW/cm2, and 30 mW/cm2, respectively, but with the same total dose (5.4 J/cm2). A control (CO) group was not exposed to ultraviolet-A. No surgery was done on the contralateral eyes. The corneas of each group were evaluated by the effective elastic modulus (Eeff) and the hydraulic conductivity (K) within a 7.5-mm radius using nanoindentation measurements. Results Compared with the CO group, Eeff (in regions with radii of 0-1.5 mm, 1.5-3.0 mm, and 3.0-4.5 mm) significantly increased by 309%, 276%, and 226%, respectively, with SCXL; by 222%, 209%, and 173%, respectively, with ACXL1; by 111%, 109%, and 94%, respectively, with ACXL2; and by 59%, 41%, and 37%, respectively, with ACXL3 (all P < 0.05). K was also significantly reduced by 84%, 81%, and 78%, respectively, with SCXL; by 75%, 74%, and 70%, respectively, with ACXL1; by 64%, 62%, and 61%, respectively, with ACXL2; and by 33%, 36%, and 32%, respectively, with ACXL3 (all P < 0.05). For the other regions(with radii between 4.5 and 7.5 mm), the SCXL and ACXL1 groups (but not the ACXL2 and ACXL3 groups) still showed significant changes in Eeff and K. Conclusions CXL had a significant effect on corneal biomechanics in both standard and accelerated procedures that may go beyond the irradiated area. The effect of CXL in stiffening the tissue and reducing permeability consistently decreased with reducing the irradiance duration.
Collapse
Affiliation(s)
- Xiaobo Zheng
- School of Aeronautics, Northwestern Polytechnical University, Xi'an, China
- National Engineering Research Center of Ophthalmology and Optometry, Eye Hospital, Wenzhou Medical University, Wenzhou, China
- State Key Laboratory of Ophthalmology, Optometry and Visual Science, Eye Hospital, Wenzhou Medical University, Wenzhou, China
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
| | - Yue Xin
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
- Dalian Medical University, Affiliated Dalian No. 3 People's Hospital, Dalian, China
| | - Chong Wang
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
| | - Yiwen Fan
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
| | - Peng Yang
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
| | - Lingqiao Li
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
| | - Danping Yin
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
| | - Erchi Zhang
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
| | - Yuxin Hong
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
| | - Han Bao
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
| | - Junjie Wang
- National Engineering Research Center of Ophthalmology and Optometry, Eye Hospital, Wenzhou Medical University, Wenzhou, China
- State Key Laboratory of Ophthalmology, Optometry and Visual Science, Eye Hospital, Wenzhou Medical University, Wenzhou, China
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
| | - Fangjun Bao
- National Engineering Research Center of Ophthalmology and Optometry, Eye Hospital, Wenzhou Medical University, Wenzhou, China
- State Key Laboratory of Ophthalmology, Optometry and Visual Science, Eye Hospital, Wenzhou Medical University, Wenzhou, China
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
| | - Weiwei Zhang
- School of Aeronautics, Northwestern Polytechnical University, Xi'an, China
| | - Shihao Chen
- National Engineering Research Center of Ophthalmology and Optometry, Eye Hospital, Wenzhou Medical University, Wenzhou, China
- State Key Laboratory of Ophthalmology, Optometry and Visual Science, Eye Hospital, Wenzhou Medical University, Wenzhou, China
- National Clinical Research Center for Ocular Diseases, Eye Hospital, Wenzhou Medical University, Wenzhou, China
| | - Ahmed Elsheikh
- School of Engineering, University of Liverpool, Liverpool, United Kingdom
| | - Michael Swain
- AMME, Biomechanics Engineering, The University of Sydney, Sydney, Australia
| |
Collapse
|
3
|
Zheng X, Weng Y, Wang Y, Xin Y, Wu J, Ziad Masoud Abu Said A, Nguelemo Mayopa K, Akiti S, Li X, Wang C, Wang J, Eliasy A, Bao F, Chen S, Elsheikh A. Long-term Effects of Riboflavin Ultraviolet-A-Induced CXL With Different Irradiances on the Biomechanics of In Vivo Rabbit Corneas. J Refract Surg 2022; 38:389-397. [PMID: 35686711 DOI: 10.3928/1081597x-20220425-01] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/20/2022]
Abstract
PURPOSE To evaluate the long-term effects of ultraviolet-A corneal cross-linking (CXL) with different irrandiances on the biomechanical properties of rabbit corneas and the corresponding changes in stromal microstructure. METHODS The study involved the left eyes of 85 healthy white Japanese rabbits, randomly divided into five groups (n = 16 to 18 each). After removing the epithelium, the first four groups were exposed to riboflavin (0.22% concentration by volume) and ultraviolet-A (370 nm) at different CXL irradiations but with the same total dose (5.4 J/cm2). The four groups were defined as standard CXL (SCXL; 3 mW/cm2 for 30 minutes, n = 17), accelerated CXL1 (ACXL1; 9 mW/cm2 for 10 minutes, n = 16), accelerated CXL2 (ACXL2; 18 mW/cm2 for 5 minutes, n = 17), and accelerated CXL3 (ACXL3; 30 mW/cm2 for 3 minutes, n = 17). The control group (n = 18) was treated with riboflavin without ultraviolet-A exposure. Nine months after CXL, 10 corneas from each group were tested ex vivo under inflation, and the tangent modulus (Et) was estimated using an inverse analysis process. The remaining six to eight specimens in each group were examined by electron microscopy to determine the mean fibril diameter and interfibrillar spacing. RESULTS The SCXL and ACXL1 groups showed statistically significant differences in Et at all stresses (0.005, 0.010, and 0.015 MPa) analyzed compared to the control group (all P < .01), but the differences were non-significant in the ACXL3 group (P = 1.000, .785, and .679, respectively). For the ACXL2 group, there was no statistical difference in Et under the low stress of 0.005 MPa (P = .155), but the differences became significant at 0.010 and 0.015 MPa when compared with the control group (all P < .05). CONCLUSIONS CXL had a significant effect on corneal biomechanics in both standard and accelerated procedures. However, standard CXL was the most effective, and this effectiveness decreased gradually with increasing ultraviolet-A power intensity. [J Refract Surg. 2022;38(6):389-397.].
Collapse
|
4
|
Quantitative interocular comparison of total corneal surface area and corneal diameter in patients with highly asymmetric keratoconus. Sci Rep 2022; 12:4276. [PMID: 35277548 PMCID: PMC8917212 DOI: 10.1038/s41598-022-08021-6] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/01/2021] [Accepted: 02/18/2022] [Indexed: 12/02/2022] Open
Abstract
Keratoconus is a progressive corneal disorder which is frequently asymmetric. The aetiology of keratoconus remains unclear, and the concept of keratoconus as an ectatic disorder has been challenged recently. We carried out a retrospective study in 160 eyes of 80 patients, to evaluate and compare interocular differences in corneal diameter and surface area in patients with unilateral or highly asymmetric keratoconus (UHAKC). Calculations were performed using raw topographic elevation data derived from topographic measurements using Orbscan II, and we extrapolated surface areas up to measured corneal diameter. We also evaluated inter-eye correlation, and correlation between corneal surface area, corneal diameter and keratoconus severity. Our results showed a statistically significant but not clinically important greater corneal diameter (12.14 mm and 12.17 mm; p = 0.04), and corneal surface area (paired t-test, p < 0.0001; p = 0.0009 respectively) in more affected eyes. Inter-eye comparison revealed corneal diameter, anterior chamber depth, and corneal surface area were strongly correlated between eyes. Corneal surface area remained strongly correlated, and Bland–Altman analysis also showed strong inter-ocular agreement. Our results show that in patients with UHAKC the interocular difference in corneal diameter and corneal surface area is clinically insignificant, and are consistent with a redistribution, rather than increase, of corneal surface area with keratoconus progression.
Collapse
|
5
|
Interocular Symmetry Analysis of Corneal Elevation Using the Fellow Eye as the Reference Surface and Machine Learning. Healthcare (Basel) 2021; 9:healthcare9121738. [PMID: 34946464 PMCID: PMC8702115 DOI: 10.3390/healthcare9121738] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/30/2021] [Revised: 12/12/2021] [Accepted: 12/13/2021] [Indexed: 11/16/2022] Open
Abstract
Unilateral corneal indices and topography maps are routinely used in practice, however, although there is consensus that fellow-eye asymmetry can be clinically significant, symmetry studies are limited to local curvature and single-point thickness or elevation measures. To improve our current practices, there is a need to devise algorithms for generating symmetry colormaps, study and categorize their patterns, and develop reference ranges for new global discriminative indices for identifying abnormal corneas. In this work, we test the feasibility of using the fellow eye as the reference surface for studying elevation symmetry throughout the entire corneal surface using 9230 raw Pentacam files from a population-based cohort of 4613 middle-aged adults. The 140 × 140 matrix of anterior elevation data in these files were handled with Python to subtract matrices, create color-coded maps, and engineer features for machine learning. The most common pattern was a monochrome circle (“flat”) denoting excellent mirror symmetry. Other discernible patterns were named “tilt”, “cone”, and “four-leaf”. Clustering was done with different combinations of features and various algorithms using Waikato Environment for Knowledge Analysis (WEKA). Our proposed approach can identify cases that may appear normal in each eye individually but need further testing. This work will be enhanced by including data of posterior elevation, thickness, and common diagnostic indices.
Collapse
|