Biomechanics of the Healthy and Keratoconic Corneas: A Combination of the Clinical Data, Finite Element Analysis, and Artificial Neural Network

Ex vivo, in vivo and in silico studies of corneal biomechanics: a systematic review

Healthy cornea guarantees the refractive power of the eye and the protection of the inner components, but injury, trauma or pathology may impair the tissue shape and/or structural organization and therefore its material properties, compromising its functionality in the ocular visual process. It turns out that biomechanical research assumes an essential role in analysing the morphology and biomechanical response of the cornea, preventing pathology occurrence, and improving/optimising treatments. In this review, ex vivo, in vivo and in silico methods for the corneal mechanical characterization are reported. Experimental techniques are distinct in testing mode (e.g., tensile, inflation tests), samples' species (human or animal), shape and condition (e.g., healthy, treated), preservation methods, setup and test protocol (e.g., preconditioning, strain rate). The meaningful results reported in the pertinent literature are discussed, analysing differences, key features and weaknesses of the methodologies adopted. In addition, numerical techniques based on the finite element method are reported, incorporating the essential steps for the development of corneal models, such as geometry, material characterization and boundary conditions, and their application in the research field to extend the experimental results by including further relevant aspects and in the clinical field for diagnostic procedure, treatment and planning surgery. This review aims to analyse the state-of-art of the bioengineering techniques developed over the years to study the corneal biomechanics, highlighting their potentiality to improve diagnosis, treatment and healing process of the corneal tissue, and, at the same, pointing out the current limits in the experimental equipment and numerical tools that are not able to fully characterize in vivo corneal tissues non-invasively and discourage the use of finite element models in daily clinical practice for surgical planning.

CorNet: Autonomous feature learning in raw Corvis ST data for keratoconus diagnosis via residual CNN approach

PURPOSE

To ascertain whether the integration of raw Corvis ST data with an end-to-end CNN can enhance the diagnosis of keratoconus (KC).

METHOD

The Corvis ST is a non-contact device for in vivo measurement of corneal biomechanics. The CorNet was trained and validated on a dataset consisting of 1786 Corvis ST raw data from 1112 normal eyes and 674 KC eyes. Each raw data consists of the anterior and posterior corneal surface elevation during air-puff induced dynamic deformation. The architecture of CorNet utilizes four ResNet-inspired convolutional structures that employ 1 × 1 convolution in identity mapping. Gradient-weighted Class Activation Mapping (Grad-CAM) was adopted to visualize the attention allocation to diagnostic areas. Discriminative performance was assessed using metrics including the AUC of ROC curve, sensitivity, specificity, precision, accuracy, and F1 score.

RESULTS

CorNet demonstrated outstanding performance in distinguishing KC from normal eyes, achieving an AUC of 0.971 (sensitivity: 92.49%, specificity: 91.54%) in the validation set, outperforming the best existing Corvis ST parameters, namely the Corvis Biomechanical Index (CBI) with an AUC of 0.947, and its updated version for Chinese populations (cCBI) with an AUC of 0.963. Though the ROC curve analysis showed no significant difference between CorNet and cCBI (p = 0.295), it indicated a notable difference between CorNet and CBI (p = 0.011). The Grad-CAM visualizations highlighted the significance of corneal deformation data during the loading phase rather than the unloading phase for KC diagnosis.

CONCLUSION

This study proposed an end-to-end CNN approach utilizing raw biomechanical data by Corvis ST for KC detection, showing effectiveness comparable to or surpassing existing parameters provided by Corvis ST. The CorNet, autonomously learning comprehensive temporal and spatial features, demonstrated a promising performance for advancing KC diagnosis in ophthalmology.

Comparison of different corneal imaging modalities using artificial intelligence for diagnosis of keratoconus: a systematic review and meta-analysis

PURPOSE

This review was designed to compare different corneal imaging modalities using artificial intelligence (AI) for the diagnosis of keratoconus (KCN), subclinical KCN (SKCN), and forme fruste KCN (FFKCN).

METHODS

A comprehensive systematic search was conducted in scientific databases, including Web of Science, PubMed, Scopus, and Google Scholar based on the PRISMA statement. Two independent reviewers assessed all potential publications on AI and KCN up to March 2022. The Critical Appraisal Skills Program (CASP) 11-item checklist was used to evaluate the validity of the studies. Eligible articles were categorized into three groups (KCN, SKCN, and FFKCN) and included in the meta-analysis. The pooled estimate of accuracy (PEA) was calculated for all selected articles.

RESULTS

The initial search yielded 575 relevant publications, of which 36 met the CASP quality criteria and were included in the analysis. Qualitative assessment showed that Scheimpflug and Placido combined with biomechanical and wavefront evaluations improved KCN detection (PEA, 99.2, and 99.0, respectively). The Scheimpflug system (92.25 PEA, 95% CI, 94.76-97.51) and a combination of Scheimpflug and Placido (96.44 PEA, 95% CI, 93.13-98.19) had the highest diagnostic accuracy for the detection of SKCN and FFKCN, respectively. The meta-analysis outcomes showed no significant difference between the CASP score and accuracy of the publications (all P > 0.05).

CONCLUSIONS

Simultaneous Scheimpflug and Placido corneal imaging methods provide high diagnostic accuracy for early detection of keratoconus. The use of AI models improves the discrimination of keratoconic eyes from normal corneas.

Corneal biomechanics in early diagnosis of keratoconus using artificial intelligence

Keratoconus is a blinding eye disease that affects activities of daily living; therefore, early diagnosis is crucial. Great efforts have been made toward an early diagnosis of keratoconus. Recent studies have shown that corneal biomechanics is associated with the occurrence and progression of keratoconus. Hence, detecting changes in corneal biomechanics may provide a novel strategy for early diagnosis. However, an early keratoconus diagnosis remains challenging due to the subtle and localized nature of its lesions. Artificial intelligence has been used to help address this problem. Herein, we reviewed the literature regarding three aspects of keratoconus (keratoconus, early keratoconus, and keratoconus grading) based on corneal biomechanical properties using artificial intelligence. Furthermore, we summarized the current research progress, limitations, and possible prospects.

Keratoconus Diagnosis: From Fundamentals to Artificial Intelligence: A Systematic Narrative Review

The remarkable recent advances in managing keratoconus, the most common corneal ectasia, encouraged researchers to conduct further studies on the disease. Despite the abundance of information about keratoconus, debates persist regarding the detection of mild cases. Early detection plays a crucial role in facilitating less invasive treatments. This review encompasses corneal data ranging from the basic sciences to the application of artificial intelligence in keratoconus patients. Diagnostic systems utilize automated decision trees, support vector machines, and various types of neural networks, incorporating input from various corneal imaging equipment. Although the integration of artificial intelligence techniques into corneal imaging devices may take time, their popularity in clinical practice is increasing. Most of the studies reviewed herein demonstrate a high discriminatory power between normal and keratoconus cases, with a relatively lower discriminatory power for subclinical keratoconus.

Biomechanical analysis of ocular diseases and its in vitro study methods

Ocular diseases are closely related to the physiological changes in the eye sphere and its contents. Using biomechanical methods to explore the relationship between the structure and function of ocular tissue is beneficial to reveal the pathological processes. Studying the pathogenesis of various ocular diseases will be helpful for the diagnosis and treatment of ocular diseases. We provide a critical review of recent biomechanical analysis of ocular diseases including glaucoma, high myopia, and diabetes. And try to summarize the research about the biomechanical changes in ocular tissues (e.g., optic nerve head, sclera, cornea, etc.) associated with those diseases. The methods of ocular biomechanics research in vitro in recent years are also reviewed, including the measurement of biomechanics by ophthalmic equipment, finite element modeling, and biomechanical analysis methods. And the preparation and application of microfluidic eye chips that emerged in recent years were summarized. It provides new inspiration and opportunity for the pathogenesis of eye diseases and personalized and precise treatment.

Computational Modeling of Ophthalmic Procedures: Computational Modeling of Ophthalmic Procedures

PURPOSE

To explore how finite-element calculations can continue to contribute to diverse problems in ophthalmology and vision science, we describe our recent work on modeling the force on the peripheral retina in intravitreal injections and how that force increases with shorter, smaller gauge needles. We also present a calculation that determines the location and stress on a retinal pigment epithelial detachment during an intravitreal injection, the possibility that stress induced by the injection can lead to a tear of the retinal pigment epithelium.

BACKGROUND

Advanced computational models can provide a critical insight into the underlying physics in many surgical procedures, which may not be intuitive.

METHODS

The simulations were implemented using COMSOL Multiphysics. We compared the monkey retinal adhesive force of 18 Pa with the results of this study to quantify the maximum retinal stress that occurs during intravitreal injections.

CONCLUSIONS

Currently used 30-gauge needles produce stress on the retina during intravitreal injections that is only slightly below the limit that can create retinal tears. As retina specialists attempt to use smaller needles, the risk of complications may increase. In addition, we find that during an intravitreal injection, the stress on the retina in a pigment epithelial detachment occurs at the edge of the detachment (found clinically), and the stress is sufficient to tear the retina. These findings may guide physicians in future clinical research. NOTE: Publication of this article is sponsored by the American Ophthalmological Society.

Single-cell atlas of keratoconus corneas revealed aberrant transcriptional signatures and implicated mechanical stretch as a trigger for keratoconus pathogenesis

Keratoconus is a common ectatic corneal disorder in adolescents and young adults that can lead to progressive visual impairment or even legal blindness. Despite the high prevalence, its etiology is not fully understood. In this study, we performed single-cell RNA sequencing (scRNA-Seq) analysis on 39,214 cells from central corneas of patients with keratoconus and healthy individuals, to define the involvement of each cell type during disease progression. We confirmed the central role of corneal stromal cells in this disease, where dysregulation of collagen and extracellular matrix (ECM) occurred. Differential gene expression and histological analyses revealed two potential novel markers for keratoconus stromal cells, namely CTSD and CTSK. Intriguingly, we detected elevated levels of YAP1 and TEAD1, the master regulators of biomechanical homeostasis, in keratoconus stromal cells. Cyclical mechanical experiments implicated the mechanical stretch in prompting protease production in corneal stromal cells during keratoconus progression. In the epithelial cells of keratoconus corneas, we observed reduced basal cells and abnormally differentiated superficial cells, unraveling the corneal epithelial lesions that were usually neglected in clinical diagnosis. In addition, several elevated cytokines in immune cells of keratoconus samples supported the involvement of inflammatory response in the progression of keratoconus. Finally, we revealed the dysregulated cell-cell communications in keratoconus, and found that only few ligand-receptor interactions were gained but a large fraction of interactional pairs was erased in keratoconus, especially for those related to protease inhibition and anti-inflammatory process. Taken together, this study facilitates the understanding of molecular mechanisms underlying keratoconus pathogenesis, providing insights into keratoconus diagnosis and potential interventions.

Anterior pituitary, sex hormones, and keratoconus: Beyond traditional targets

"The Diseases of the Horny-coat of The Eye", known today as keratoconus, is a progressive, multifactorial, non-inflammatory ectatic corneal disorder that is characterized by steepening (bulging) and thinning of the cornea, irregular astigmatism, myopia, and scarring that can cause devastating vision loss. The significant socioeconomic impact of the disease is immeasurable, as patients with keratoconus can have difficulties securing certain jobs or even joining the military. Despite the introduction of corneal crosslinking and improvements in scleral contact lens designs, corneal transplants remain the main surgical intervention for treating keratoconus refractory to medical therapy and visual rehabilitation. To-date, the etiology and pathogenesis of keratoconus remains unclear. Research studies have increased exponentially over the years, highlighting the clinical significance and international interest in this disease. Hormonal imbalances have been linked to keratoconus, both clinically and experimentally, with both sexes affected. However, it is unclear how (molecular/cellular signaling) or when (age/disease stage(s)) those hormones affect the keratoconic cornea. Previous studies have categorized the human cornea as an extragonadal tissue, showing modulation of the gonadotropins, specifically luteinizing hormone (LH) and follicle-stimulating hormone (FSH). Studies herein provide new data (both in vitro and in vivo) to further delineate the role of hormones/gonadotropins in the keratoconus pathobiology, and propose the existence of a new axis named the Hypothalamic-Pituitary-Adrenal-Corneal (HPAC) axis.

[Keratoconus detection and classification from parameters of the Corvis®ST : A study based on algorithms of machine learning]

BACKGROUND AND OBJECTIVE

In the last decades increasingly more systems of artificial intelligence have been established in medicine, which identify diseases or pathologies or discriminate them from complimentary diseases. Up to now the Corvis®ST (Corneal Visualization Scheimpflug Technology, Corvis®ST, Oculus, Wetzlar, Germany) yielded a binary index for classifying keratoconus but did not enable staging. The purpose of this study was to develop a prediction model, which mimics the topographic keratoconus classification index (TKC) of the Pentacam high resolution (HR, Oculus) with measurement parameters extracted from the Corvis®ST.

PATIENTS AND METHODS

In this study 60 measurements from normal subjects (TKC 0) and 379 eyes with keratoconus (TKC 1-4) were recruited. After measurement with the Pentacam HR (target parameter TKC) a measurement with the Corvis®ST device was performed. From this device 6 dynamic response parameters were extracted, which were included in the Corvis biomechanical index (CBI) provided by the Corvis®ST (ARTh, SP-A1, DA ratio 1 mm, DA ratio 2 mm, A1 velocity, max. deformation amplitude). In addition to the TKC as the target, the binarized TKC (1: TKC 1-4, 0: TKC 0) was modelled. The performance of the model was validated with accuracy as an indicator for correct classification made by the algorithm. Misclassifications in the modelling were penalized by the number of stages of deviation between the modelled and measured TKC values.

RESULTS

A total of 24 different models of supervised machine learning from 6 different families were tested. For modelling of the TKC stages 0-4, the algorithm based on a support vector machine (SVM) with linear kernel showed the best performance with an accuracy of 65.1% correct classifications. For modelling of binarized TKC, a decision tree with a coarse resolution showed a superior performance with an accuracy of 95.2% correct classifications followed by the SVM with linear or quadratic kernel and a nearest neighborhood classifier with cubic kernel (94.5% each).

CONCLUSION

This study aimed to show the principle of supervised machine learning applied to a set-up for the modelled classification of keratoconus staging. Preprocessed measurement data extracted from the Corvis®ST device were used to mimic the TKC provided by the Pentacam device with a series of different algorithms of machine learning.

Association between Corneal Stiffness Parameter at the First Applanation and Keratoconus Severity

Objective

The study aimed to evaluate the character of corneal stiffness parameter at the first applanation (SP-A1) in normal and keratoconus eyes and explore the association between SP-A1 and keratoconus severity indicators.

Methods

A total of 351 normal and 351 keratoconus eyes were included in the current study. Keratoconus was diagnosed according to the corneal topography map and slit-lamp examination. The severity of keratoconus was classified to mild (steep keratometry (Ks) < 48D), moderate (48 ≤ Ks < 55D), and severe (Ks ≥ 55D). The SP-A1 was measured using the Corvis ST software. The correlation analyses and receiver operating characteristic (ROC) curve were performed in the current analysis.

Results

The SP-A1 values of keratoconus were lower than that of normal eyes (72.11 (57.02, 83.08) mmHg/mm vs 110.89 (100.45, 122.47) mmHg/mm, P < 0.001). With the severity of keratoconus increasing, the SP-A1 decreased and the value of SP-A1 was 79.54 (70.30, 90.93) mmHg/mm, 65.11 (53.14, 77.46) mmHg/mm, and 47.59 (37.50, 62.14) mmHg/mm in mild, moderate, and severe keratoconus eyes, respectively (P < 0.001). The negative association between SP-A1 and Ks was found in mild, moderate, and severe keratoconus eyes (r _mild = -0.171, r _moderate = -0.317, r _severe = -0.288, all P < 0.05). A positive association between SP-A1 and the thinnest corneal thickness (TCT) was found in all eyes (r_normal = 0.687, r _mild = 0.519, r _moderate = 0.488, r _severe = 0.382, all P < 0.05). SP-A1 was found to be statistically positively associated with intraocular pressure (IOP), biomechanical corrected IOP (bIOP), time from the initiation of air puff until the first applanation (A1T), corneal velocity at the second applanation (A2V), and negatively associated with deformation amplitude (DA), peak distance (PD), corneal velocity at the first applanation (A1V), time from the initiation of air puff until the second applanation (A2T), and DA Ratio Max [2 mm] both in normal and keratoconus eyes (all P < 0.05). The ROC analysis indicated that the AUC (95% CI) of SP-A1 was 0.952 (0.934-0.967) and 0.930 (0.904-0.951) in detecting keratoconus eyes and mild keratoconus eyes from normal eyes, respectively.

Conclusions

The SP-A1 value decreased while the keratoconus severity increased. It was lower in keratoconus than that in normal eyes and could be helpful in identifying keratoconus eyes from normal eyes. Further researches would be warranted to expand the clinical utility of SP-A1.

Biomechanics of the keratoconic cornea: Theory, segmentation, pressure distribution, and coupled FE-optimization algorithm

Understanding of the corneal biomechanical properties is of high interest due to its potential application in the early diagnosis of keratoconus (KC). KC by itself is a non-inflammatory eye disorder causes corneal structural and/or compositional anomalies. The biomechanically weakened cornea is no longer able to preserve the normal shape of the cornea against the intraocular pressure (IOP) and gradually starts to bulge outward, invoking a conical shape and subsequent distorted vision. The most popular way to measure the in vivo corneal biomechanical properties is the CorVis-ST, which enables to analyze the dynamic response of the cornea under a temporary air puff pressure. However, the complications, such as the lack of knowledge on the accurate air-puff pressure distribution on the cornea's surface as a function of the distance from the apex of the cornea as well as the time, hinder us to have a reliable estimation of the cornea's mechanical parameters. This study aims to establish patient-specific geometries of the healthy and KC corneas and calculate the pressure distribution on the cornea's surface as a function of both the distance from the apex of the cornea and time, and thereafter, the viscoelastic mechanical properties of both the healthy and KC corneas using a coupled finite element (FE)-optimization algorithm. To do that, the dynamic deformation response of six healthy and six KC corneas were measured via CorVis-ST. The videos of the in vivo deformation of the corneas under the applied air puff pressure were segmented using our segmentation algorithm to determine the anterior and posterior curvatures of the corneas during the dynamic movement of the cornea. The FE model of the corneas were established using the segmented data and subjected to a negative (pre-stress), positive IOP, and air puff pressure while the floating boundary conditions were applied to the two ends of the corneas' FE models. The simulation results were imported into a loop of FE-optimization algorithm and analyzed until the deformation amplitude at the apex of the cornea reaches to its minimum difference compared to the clinical data by CorVis-ST. The results revealed that the pressure distributions found in the literature as a function of the distance from the apex of the cornea and time unable to provide satisfactory results. Therefore, the pressure distributions both as a function of the distance and time were optimized using our coupled FE-optimization algorithm and employed to estimate the viscoelastic properties of the healthy and KC corneas. The mean percentage error (MPE) of 8.45% and 10.79% were found for the healthy and KC corneas compared to the clinical data of CorVis-ST, respectively. The results also revealed a significantly higher short-time shear modulus for the KC (62.33 MPa) compared to the healthy (37.45 MPa) corneas while the long-time shear modulus of both the healthy and KC corneas were almost the same (4.01 vs. 3.91 MPa). The proposed algorithm is a noninvasive technique capable of accurately estimating the viscoelastic mechanical properties of the cornea, which can contribute to understand the mechanism of KC development and improve diagnosis and intervention in KC.

Topographic and Biomechanical Changes after Application of Corneal Cross-Linking in Recurrent Keratoconus

Background: Recurrent keratoconus (RKC) develops as a progressive thinning of the peripheral and the inferior cornea after keratoplasty, in both graft and host, causing secondary astigmatism, refractive instability, and reduced visual acuity. We evaluated the effectiveness of corneal cross-linking (CXL) in patients diagnosed with RKC. Methods: Accelerated-CXL via the epi-off technique was performed in15 patients (18 eyes) diagnosed with RKC. Topographic and biomechanical changes were assessed at 12 months. Results: Differences in maximum keratometry, thinnest corneal thickness, and biomechanical parameters (deformation amplituderatio, inverse concave radius, applanation 1 velocity, and applanation 2 velocity, stiffness A1) versus baseline were statistically significant (p < 0.05).Best corrected visual acuity was improved in 13 eyes and unchanged in 4;manifest refractive spherical equivalent was reduced in 13 eyes, increased in 3,and unchanged in 1 eye; topographic astigmatism was reduced in 9 eyes, remained stable in 1 eye, and increased in 7 eyes. Conclusions: Improved topographic and biomechanic indexes at 1 year after CXL suggest it's potential as first-line therapy for RKC, as it is for KC.