Implementing and evaluating a fully functional AI-enabled model for chronic eye disease screening in a real clinical environment.

BACKGROUND: Artificial intelligence (AI) has the potential to increase the affordability and accessibility of eye disease screening, especially with the recent approval of AI-based diabetic retinopathy (DR) screening programs in several countries. METHODS: This study investigated the performance, feasibility, and user experience of a seamless hardware and software solution for screening chronic eye diseases in a real-world clinical environment in Germany. The solution integrated AI grading for DR, age-related macular degeneration (AMD), and glaucoma, along with specialist auditing and patient referral decision. The study comprised several components: (1) evaluating the entire system solution from recruitment to eye image capture and AI grading for DR, AMD, and glaucoma; (2) comparing specialist's grading results with AI grading results; (3) gathering user feedback on the solution. RESULTS: A total of 231 patients were recruited, and their consent forms were obtained. The sensitivity, specificity, and area under the curve for DR grading were 100.00%, 80.10%, and 90.00%, respectively. For AMD grading, the values were 90.91%, 78.79%, and 85.00%, and for glaucoma grading, the values were 93.26%, 76.76%, and 85.00%. The analysis of all false positive cases across the three diseases and their comparison with the final referral decisions revealed that only 17 patients were falsely referred among the 231 patients. The efficacy analysis of the system demonstrated the effectiveness of the AI grading process in the study's testing environment. Clinical staff involved in using the system provided positive feedback on the disease screening process, particularly praising the seamless workflow from patient registration to image transmission and obtaining the final result. Results from a questionnaire completed by 12 participants indicated that most found the system easy, quick, and highly satisfactory. The study also revealed room for improvement in the AMD model, suggesting the need to enhance its training data. Furthermore, the performance of the glaucoma model grading could be improved by incorporating additional measures such as intraocular pressure. CONCLUSIONS: The implementation of the AI-based approach for screening three chronic eye diseases proved effective in real-world settings, earning positive feedback on the usability of the integrated platform from both the screening staff and auditors. The auditing function has proven valuable for obtaining efficient second opinions from experts, pointing to its potential for enhancing remote screening capabilities. TRIAL REGISTRATION: Institutional Review Board of the Hamburg Medical Chamber (Ethik-Kommission der Ärztekammer Hamburg): 2021-10574-BO-ff.

Quality of life in stable and progressive 'early-stage' keratoconus patients.

PURPOSE: To analyse the vision-related quality of life (vr-QoL) in stable and progressive keratoconus (KC) patients with a still good visual acuity. METHODS: Combined prospective/cross-sectional study design. The Refractive Status and Vision Profile (RSVP) and the National Eye Institute Visual Functioning - 25 (NEI-25) questionnaire were used in 16 emmetropic, 32 myopic and 56 KC patients, whereby KC patients with a stable (n = 26) and patients with a progressive stage (n = 30) and some of them before and after corneal cross-linking (CXL; n = 10) were included. All patients had a DCVA in at least one eye of ≥0.7 (decimal chart). RESULTS: The analyses revealed a minor decline of the vr-QoL from emmetropes to myopes to early KC patients with a stable disease. Nonetheless, sociological subscales (i.e. 'social functioning', 'role difficulties' and 'dependency') did not display statistically significant differences comparing these groups. In progressive KC, we could demonstrate a statistically significant decline also of these sociological subscales, which did not improve after CXL. CONCLUSION: Due to a still high and almost unaffected vr-QoL in early KC patients and the distinct decline after progression without rehabilitation after CXL, a reconsideration of current strategies to perform CXL only after a progression is diagnosed should be re-evaluated.

In vivo Histology of the Cornea - from the "Rostock Cornea Module" to the "Rostock Electronic Slit Lamp" - a Clinical "Proof of Concept" Study. / In-vivo-Histologie der Hornhaut ­ vom "Rostocker Cornea Modul" zur "Rostocker Elektronischen Spaltlampe" ­ eine klinische "Proof-of-Concept"-Studie.

INTRODUCTION: Confocal in vivo microscopy is an established method in ophthalmology research. As it requires contact coupling and calibration of the instruments is suboptimal, this method has been only rarely used in clinical routine work. As a result of close collaboration between physicists, information scientists and ophthalmologists, confocal laser scanning microscopy (CLSM) of the eye has been developed in recent years and a prototype can now be used in patients. The present study evaluates possible clinical uses of this method. MATERIAL AND METHODS: The essential innovations in CLSM are (1) a newly designed coupling element with superficial adaptation to corneal curvature and (2) the use of a dual computerised piezo drive for rapid and precise focusing. In post-processing and after elastic imaging registration of the individual images parallel to the surface, it is also possible to produce sagittal sections resembling a split lamp and with resolution in the micrometer range. The concept was tested on enucleated pig bulbi and tested on normal volunteers and selected patients with diseases of the cornea. RESULTS: Simultaneous imaging in planes parallel to the surface and in sagittal planes provided additional information that can help us to understand the processes of wound healing in all substructures of the cornea and the role of immune competent cells. Possible clinical uses were demonstrated in a volunteer with healthy eyes and several groups of patients (keratoconus after CXL, recurrent keratitis, status after PRK). These show that this new approach can be used in morphological studies at cellular level in any desired and appropriate test plane. CONCLUSIONS: It could be shown that this new concept of CLSM can be used clinically. It can provide valuable and novel information to both preclinical researchers and to ophthalmologists interested in corneal disease, e.g. density of Langerhans cells and epithelial stratification in ocular surface diseases.

Whole-Exome Sequencing Identifies Novel Compound Heterozygous ZNF469 Mutations in Two Siblings with Mild Brittle Cornea Syndrome.

Connective tissue diseases, including osteogenesis imperfecta (OI) and Ehlers-Danlos syndrome (EDS), exhibit a high degree of clinical and genetic heterogeneity. We report two sisters with blue sclerae, joint hypermobility and hearing loss. Whole-exome sequencing identified two compound heterozygous ZNF469 loss-of-function mutations due to a frameshift. Since these findings indicate the presence of brittle cornea syndrome (BCS), we performed ocular optical coherence tomography (OCT) and pachymetry, which revealed a moderate decrease in corneal thickness. While only one traumatic fracture was observed in each of the patients, a detailed skeletal assessment indicated no specific patterns of bone mass and microstructure reduction as well as normal bone turnover markers. Taken together, our findings point to a mild form of brittle cornea syndrome with a phenotype compatible with the extraskeletal features of OI but also with EDS.

Features of Marfan syndrome not listed in the Ghent nosology - the dark side of the disease.

Introduction: The revised Ghent nosology presents the classical features of Marfan syndrome. However, behind its familiar face, Marfan syndrome hides less well-known features.Areas covered: The German Marfan Organization listed unusual symptoms and clinical experts reviewed the literature on clinical features of Marfan syndrome not listed in the Ghent nosology. Thereby we identified the following features: (1) bicuspid aortic valve, mitral valve prolapse, pulmonary valve prolapse, tricuspid valve prolapse, (2) heart failure and cardiomyopathy, (3) supraventricular arrhythmia, ventricular arrhythmia, and abnormal repolarization, (4) spontaneous coronary artery dissection, anomalous coronary arteries, and atherosclerotic coronary artery disease, tortuosity-, aneurysm-, and dissection of large and medium-sized arteries, (5) restrictive lung disease, parenchymal lung disease, and airway disorders, (6) obstructive- and central sleep apnea, (7) liver and kidney cysts, biliary tract disease, diaphragmatic hernia, and adiposity, (8) premature labor, and urinary incontinence, (9) myopathy, reduced bone mineral density, and craniofacial manifestations, (10) atrophic scars, (11) caries, and craniomandibular dysfunction, (12) headache from migraine and spontaneous cerebrospinal fluid leakage, (13) cognitive dysfunction, schizophrenia, depression, fatigue, and pain, (14) and activated fibrinolysis, thrombin, platelets, acquired von Willebrand disease, and platelet dysfunction.Expert commentary: Future research, nosologies, and guidelines may consider less well-known features of Marfan syndrome.

Analysis of excimer laser treatment outcomes and corresponding angle κ in hyperopic astigmatism.

PURPOSE: To investigate the prevalence of preoperative, intraoperative, and postoperative angle κ in hyperopic eyes and its effect on the refractive outcomes of excimer laser vision correction. SETTING: University Hospital Hamburg and Care Vision Refractive Centers, Germany. DESIGN: Retrospective multicenter study. METHODS: Hyperopic patients after laser in situ keratomileusis and before retreatment because of hyperopic regression were included. Three treatment groups were defined based on the magnitude of preoperative angle κ defined as low (<0.25 mm), moderate (0.25 to 0.50 mm), or high (>0.50 mm). RESULTS: The study analyzed 170 hyperopic eyes of 112 patients (mean age 29.8 years ± 10.04 [SD], range 21 to 62 years). The preoperative angle κ was low, moderate, or high in 23, 49, and 98 cases, respectively. In eyes with a preoperative angle κ of 0.25 mm or more, changes in the x-axes and y-axes of angle κ and its magnitude were statistically significant (P < .05) between preoperative and intraoperative measurements. In these eyes, intraoperative angle κ (offset) was statistically significantly smaller (P < .05), and Purkinje images were statistically significantly more caudal (P < .001). Treatment predictability was independent from using preoperative or intraoperative angle κ, and there were no statistically significant differences in efficacy or safety. CONCLUSIONS: A preoperative magnitude of angle κ as high as 0.50 mm or higher was present in more than 50% of regressed hyperopic eyes. Measuring angle κ preoperatively did not reflect its real-time intraoperative magnitude. Using the preoperatively larger angle κ as an intraoperative offset might result in a too nasally centered ablation that causes topographic and optical drawbacks; therefore, centering the ablation based on the intraoperative measured offset is recommended.

Tomographic and Biomechanical Scheimpflug Imaging for Keratoconus Characterization: A Validation of Current Indices.

PURPOSE: To analyze the potential benefit of the newly developed Tomography and Biomechanical Index (TBI) for early keratoconus screening. METHODS: In this retrospective study, the discriminatory power of the corneal tomography Belin/Ambrósio Enhanced Ectasia Display (BAD-D) index and the newly developed Corvis Biomechanical Index (CBI) and TBI to differentiate between normal eyes, manifest keratoconus eyes (KCE), very asymmetric keratoconus eyes with ectasia (VAE-E), and their fellow eyes with either regular topography (VAE-NT) or regular topography and tomography (VAE-NTT) were analyzed by applying the t test (for normal distribution), Wilcoxon matched-pairs test (if not normally distributed), and receiver operating characteristic curve (ROC). The DeLong test was used to compare the area under the ROC (AUROC). Further, the cut-offs of the analyzed indices presented in a study by Ambrósio et al. from 2017 were applied in the study population to enable a cross-validation in an independent study population. RESULTS: All indices demonstrated a high discriminative power when comparing normal and advanced keratoconus, which decreased when comparing normal and VAE-NT eyes and further when analyzing normal versus VAE-NTT eyes. The difference between the AUROCs reached a statistically significant level when comparing TBI versus BAD-D analyzing normal versus all included keratoconic eyes (P = .02). The TBI presented with the highest AUROCs throughout all conducted analyses when comparing different keratoconus stages, although not reaching a statistically significant level. Applying the cut-offs presented by Ambrósio et al. to differentiate between normal and VAE-NT in the study population, the accuracy was reproducible (accuracy in our study population with an optimized TBI cut-off: 0.72, with the cut-off defined by Ambrósio et al. 0.67). CONCLUSIONS: The TBI enables karatoconus screening in topographical and tomographical regular keratoconic eyes. To further improve the screening accuray, prospective studies should be conducted. [J Refract Surg. 2018;34(12):840-847.].

Endothelial Cell Count in Eye Bank Corneal Grafts: Impact of Death Cause and Donor Diseases.

PURPOSE: The purpose of this study is to analyze the impact of death causes and documented donor diseases on initial endothelial cell counts (after retrieval) and the development of corneal graft endothelia during organ culture. METHODS: The retrospective statistic analyses was conducted on a data set of 10,185 human corneas prepared at the Hamburg Eye Bank. RESULTS: Although we observed that death by gunshot trauma or alcoholism seems to be associated with marginally higher endothelium cell counts (independently from donor age), we could prove that only donor age is a relevant predictive parameter for the initial cell-density of the endothelium and its development in vitro. CONCLUSION: We conclude that an extension of prospective quality parameters for donor selection additional to donor age (such as individual causes of death) is not necessary.

Outcomes of retreatment after hyperopic laser in situ keratomileusis.

PURPOSE: To evaluate the efficacy, predictability, and safety of laser in situ keratomileusis (LASIK) retreatment in eyes with hyperopia. SETTING: Multicenter refractive surgery centers and University Hospital, Germany. DESIGN: Retrospective case series. METHODS: This multicenter study included hyperopic patients with a preoperative difference between cycloplegic and manifest refraction of 1.00 diopter (D) or less who had LASIK retreatment based on manifest refraction. The refractive outcome was analyzed according to standard graphs for reporting the efficacy, predictability, and safety of refractive surgery. RESULTS: One hundred thirteen eyes of 113 consecutive hyperopic patients were enrolled. Efficacy (P < .001) and safety (P = .004) were statistically significantly improved by retreatment without being negatively influenced by preoperative manifest spherical equivalent (SE), manifest cylinder, or keratometry. In cases still showing a trend toward undercorrection, retreatment resulted in 88 eyes (78.0%) being within ±0.50 D of the attempted correction. The optical zone (OZ) diameter of the retreatment did not correlate with efficacy, predictability, or safety. CONCLUSIONS: Retreatment after hyperopic LASIK resulted in high efficacy, predictability, and safety outcomes. The efficacy and safety of the retreatment were not affected by preoperative manifest SE, manifest cylinder, keratometry, or OZ diameter.

Keratoconus Screening With Dynamic Biomechanical In Vivo Scheimpflug Analyses: A Proof-of-Concept Study.

PURPOSE: This proof-of-concept study was designed to analyze the ability of in vivo biomechanical corneal analyses with the corneal visualization Scheimpflug technology (CorvisST; Oculus Optikgeräte, Wetzlar, Germany) to differentiate between normal eyes and eyes with manifest keratoconus after strictly eliminating the potential confounding factors intraocular pressure (IOP) and central corneal thickness (CCT). METHODS: In this retrospective, cross-sectional study, after pairwise matching for CCT and IOP, 29 normal eyes and 29 keratoconic eyes (one eye from each patient) were selected as study population. Older CorvisST parameters and the new Corvis Biomechanical Index (CBI), including several biomechanical and one tomographic parameter, as well as an adjusted CBI (aCBI) (including only biomechanical parameters), were compared regarding their discriminative ability between both groups. RESULTS: None of the CorvisST parameters of the former software version demonstrated statistically significant differences between normal and keratoconic eyes. On the other hand, the CBI and aCBI reached accuracies of 0.91 and 0.93, respectively, to discriminate between CCT- and IOP-matched normal and keratoconic eyes (CBI: [AUC/sensitivity/specificity]: 0.961/0.90/0.93; aCBI: [AUC/sensitivity/specificity]: 0.986/0.93/0.93). CONCLUSIONS: This study demonstrated that the concept of keratoconus screening with the CorvisST is effective in differentiating keratoconic from non-keratoconic eyes. The next steps will be testing the indices in subclinical keratoconus cases and hopefully combining biomechanical analyses with already established topography and tomography indices to further improve current keratoconus screening. [J Refract Surg. 2017;33(11):773-778.].

Intraoperative aberrometry-based aphakia refraction in patients with cataract: status and options.

AIM: To explore the application of intraoperative wavefront aberrometry (IWA) for aphakia-based biometry using three existing formulae derived from autorefractive retinoscopy and introducing new improved formulae. METHODS: In 74 patients undergoing cataract surgery, three repeated measurements of aphakic spherical equivalent (SE) were taken. All measurements were objectively graded for their quality and evaluated with the 'limits of agreement' approach. ORs were calculated and analysis of variance was applied. The intraocular lens (IOL) power that would have given the target refraction was back-calculated from manifest refraction at 3 months postoperatively. Regression analysis was performed to generate two aphakic SE-based formulae for predicting this IOL. The accuracy of the formulae was determined by comparing them to conventional biometry and published aphakia formulae. RESULTS: In 32 eyes, three consecutive aphakic measurements were successful. Objective parameters of IWA map quality significantly impacted measurement variability (p<0.05). The limits of agreement of repeated aphakic SE readings were +0.66 dioptre (D) and -0.69 D. Intraoperative biometry by our formula resulted in 25% and 53% of all cases ±0.50D and ±1.00 D within SE target, respectively. A second formula that took axial length (AL) into account resulted in improved ratios of 41% and 70%, respectively. CONCLUSIONS: A reliable application of IWA to calculate IOL power during routine cataract surgery may not be feasible given the high rate of measurement failures and the large variations of the readings. To enable reliable IOL calculation from IWA, measurement precision must be improved and aphakic IOL formulae need to be fine-tuned.

Pseudolentogenic astigmatic effect of multifocal intraocular lenses: non-corneal ocular residual astigmatism (N-CORA) as a new parameter in astigmatic change analysis.

PURPOSE: This study was initiated to introduce the term non-corneal ocular residual astigmatism (N-CORA) as a new parameter in astigmatic change analysis after implantation of two different types of non-toric, multifocal intraocular lenses (MIOL). METHODS: Seventy-two eyes from 72 consecutive patients after MIOL surgery were studied in terms of a retrospective, cross-sectional data analysis. Two types of spherical MIOL were used. Surgical technique in all patients was a 2.4-mm incision phacoemulsification, performed by one surgeon. To investigate the magnitude and axis of astigmatic changes, the true corneal astigmatism and Alpins vector method were applied. RESULTS: There were no statistically significant between-group differences related to the preoperative refraction or ocular residual astigmatism (ORA). After surgery, the mean refractive surgically induced astigmatism (RSIA) and the topographic SIA (TSIA) did not differ significantly between the lenses. The magnitude and orientation of ORA and N-CORA changed after surgery. CONCLUSIONS: There are no statistically significant differences in postoperative ORA in magnitude or axis when implanting different types of MIOL. The similarity of N-CORA between both MIOL types shows that both diffractive and refractive asymmetric MIOLs with plate haptics have the same pseudolentogenic astigmatic effect which could be presented in terms of the newly introduced parameter N-CORA.

Ocular residual astigmatism (ORA) in pre-cataract eyes prior to and after refractive lens exchange.

The purpose of this study was to analyze ocular residual astigmatism (ORA) before and after implantation of two different optical types of non-toric multifocal intraocular lenses (MIOL) in pre-cataract patients. This retrospective cohort study analyzed 72 eyes from 72 consecutive patients after MIOL surgery . To investigate magnitude and axis of astigmatic changes, the concepts of true corneal astigmatism and Alpins vector method were applied. There were no statistically significant between-group differences prior to surgery. The mean refractive surgically induced astigmatism (RSIA) (P = 0.063) and the topographic SIA (TSIA) (P = 0.828) did not differ significantly between the lenses, and the summated vector mean for ORA was reduced in terms of magnitude by approximately 0.30 Diopter. ORA in pseudophakic eyes mainly results from the posterior corneal surface and less from IOL tilting, postoperative posterior capsule shrinkage, or secondary cataract.

Semi-quantitative assessments of dextran toxicity on corneal endothelium: conceptual design of a predictive algorithm.

Dextran is added to corneal culture medium for at least 8 h prior to transplantation to ensure that the cornea is osmotically dehydrated. It is presumed that dextran has a certain toxic effect on corneal endothelium but the degree and the kinetics of this effect have not been quantified so far. We consider that such data regarding the toxicity of dextran on the corneal endothelium could have an impact on scheduling and logistics of corneal preparation in eye banking. In retrospective statistic analyses, we compared the progress of corneal endothelium (endothelium cell loss per day) of 1334 organ-cultured corneal explants in media with and without dextran. Also, the influence of donor-age, sex and cause of death on the observed dextran-mediated effect on endothelial cell counts was studied. Corneas cultured in dextran-free medium showed a mean endothelium cell count decrease of 0.7% per day. Dextran supplementation led to a mean endothelium cell loss of 2.01% per day; this reflects an increase by the factor of 2.9. The toxic impact of dextran was found to be time dependent; while the prevailing part of the effect was observed within the first 24 h after dextran-addition. Donor age, sex and cause of death did not seem to have an influence on the dextran-mediated toxicity. Based on these findings, we could design an algorithm which approximately describes the kinetics of dextran-toxicity. We reproduced the previously reported toxic effect of dextran on the corneal endothelium in vitro. Additionally, this is the first work that provides an algorithmic instrument for the semi-quantitative calculation of the putative endothelium cell count decrease in dextran containing medium for a given incubation time and could thus influence the time management and planning of corneal transplantations.

The role of the multidisciplinary health care team in the management of patients with Marfan syndrome.

Marfan syndrome (MFS) is a rare, severe, chronic, life-threatening disease with multiorgan involvement that requires optimal multidisciplinary care to normalize both prognosis and quality of life. In this article, each key team member of all the medical disciplines of a multidisciplinary health care team at the Hamburg Marfan center gives a personal account of his or her contribution in the management of patients with MFS. The authors show how, with the support of health care managers, key team members organize themselves in an organizational structure to create a common meaning, to maximize therapeutic success for patients with MFS. First, we show how the initiative and collaboration of patient representatives, scientists, and physicians resulted in the foundation of Marfan centers, initially in the US and later in Germany, and how and why such centers evolved over time. Then, we elucidate the three main structural elements; a team of coordinators, core disciplines, and auxiliary disciplines of health care. Moreover, we explain how a multidisciplinary health care team integrates into many other health care structures of a university medical center, including external quality assurance; quality management system; clinical risk management; center for rare diseases; aorta center; health care teams for pregnancy, for neonates, and for rehabilitation; and in structures for patient centeredness. We provide accounts of medical goals and standards for each core discipline, including pediatricians, pediatric cardiologists, cardiologists, human geneticists, heart surgeons, vascular surgeons, vascular interventionists, orthopedic surgeons, ophthalmologists, and nurses; and of auxiliary disciplines including forensic pathologists, radiologists, rhythmologists, pulmonologists, sleep specialists, orthodontists, dentists, neurologists, obstetric surgeons, psychiatrist/psychologist, and rehabilitation specialists. We conclude that a multidisciplinary health care team is a means to maximize therapeutic success.

Role of preoperative cycloplegic refraction in LASIK treatment of hyperopia.

BACKGROUND: Previous studies have suggested that, to improve refractive predictability in hyperopic LASIK treatments, preoperative cycloplegic or manifest refraction, or a combination of both, could be used in the laser nomogram. We set out to investigate (1) the prevalence of a high difference between manifest and cycloplegic spherical equivalent in hyperopic eyes preoperatively, and (2) the related predictability of postoperative keratometry. METHODS: Retrospective cross-sectional data analysis of consecutive treated 186 eyes from 186 consecutive hyperopic patients (mean age 42 [±12] years) were analyzed. Excimer ablation for all eyes was performed using a mechanical microkeratome (SBK, Moria, France) and an Allegretto excimer laser platform. Two groups were defined according to the difference between manifest and cycloplegic spherical equivalent which was defined as ≥1.00 diopter (D); the data was analyzed according to refractive outcome in terms of refractive predictability, efficacy, and safety. RESULTS: In 24 eyes (13 %), a preoperative difference of ≥1.00D between manifest spherical equivalent and cycloplegic spherical equivalent (= MCD) occurred. With increasing preoperative MCD, the postoperative achieved spherical equivalent showed hyperopic regression after 3 months. There was no statistically significant effect of age (accommodation) or optical zone size on the achieved spherical equivalent. CONCLUSIONS: A difference of ≥1.00D occurs in about 13 % of hyperopia cases. We suggest that hyperopic correction should be based on the manifest spherical equivalent in eyes with preoperative MCD <1.00D. If the preoperative MCD is ≥1.00D, treatment may produce manifest undercorrection, and therefore we advise that the patient should be warrned about lower predictability, and suggest basing conclusions on the arithmetic mean calculated from the preoperative manifest and cycloplegic spheres.