Open Access
Cornea & External Disease  |   April 2025
A Novel Approach to Fabricate Early Keratoconus Phantom Models
Author Affiliations & Notes
  • Hui Tong
    School of Biomedical Engineering, Capital Medical University, Beijing, China
    Beijing Key Laboratory of Fundamental Research on Biomechanics in Clinical Application, Capital Medical University, Beijing, China
  • Mingjue Wu
    School of Biomedical Engineering, Capital Medical University, Beijing, China
  • Jianqiang Han
    School of Biomedical Engineering, Capital Medical University, Beijing, China
  • Lin Li
    School of Biomedical Engineering, Capital Medical University, Beijing, China
    Beijing Key Laboratory of Fundamental Research on Biomechanics in Clinical Application, Capital Medical University, Beijing, China
  • Haixia Zhang
    School of Biomedical Engineering, Capital Medical University, Beijing, China
    Beijing Key Laboratory of Fundamental Research on Biomechanics in Clinical Application, Capital Medical University, Beijing, China
  • Correspondence: Haixia Zhang, School of Biomedical Engineering, Capital Medical University, No. 10 Xitoutiao, Youanmenwai Street, Fengtai District, Beijing 100069, China. e-mail: [email protected] 
Translational Vision Science & Technology April 2025, Vol.14, 18. doi:https://doi.org/10.1167/tvst.14.4.18
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Hui Tong, Mingjue Wu, Jianqiang Han, Lin Li, Haixia Zhang; A Novel Approach to Fabricate Early Keratoconus Phantom Models. Trans. Vis. Sci. Tech. 2025;14(4):18. https://doi.org/10.1167/tvst.14.4.18.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: To develop a method to fabricate early keratoconus phantom models and evaluate the feasibility of using corneal models for studying the dynamic response of early keratoconus under an air puff.

Methods: A corneal mold was designed, and the silicone material was poured into the mold to produce corneal phantoms. Two types of early keratoconus phantoms with reduced mechanical properties in a specific area were prepared using a two-step molding process: the central keratoconus phantom and the paracentral keratoconus phantom. Corneal Visualization Scheimpflug Technology tonometry was performed on the normal corneal phantoms and early keratoconus phantoms, and the corresponding dynamic corneal response (DCR) parameters were recorded.

Results: A majority of DCR parameters of the normal corneal phantoms, including deflection amplitude at highest concavity (HCDA), peak distance (PD), radius of curvature (HR), first and second applanation times (A1T and A2T), first and second applanation velocities (A1V and A2V), and the stiffness parameter at the first applanation (SPA1), exhibited trends in response to changes in the simulated intraocular pressure (SIOP) that aligned with experimental results based on ex vivo animal eyes. Significant differences in HCDA, PD, HR, A1V, A2V, A1T, A2T, and integrated radius (IR) were observed between the early keratoconus phantoms and the normal corneal phantoms.

Conclusions: The early keratoconus phantom models fabricated by the present novel approach are feasible for studying the dynamic response of early keratoconus under an air puff.

Translational Relevance: This study demonstrated the potential of corneal phantom models for corneal biomechanical studies, which can deepen our understanding of the DCR parameters, and the results will provide valuable information for early diagnosis of keratoconus.

Introduction
Early diagnosis of keratoconus is essential to early intervention of keratoconus and preoperative screening of corneal refractive surgery.1 In the early stages of keratoconus, the morphology of the cornea has not yet undergone significant changes, such as the forme fruste keratoconus.2,3 Studies have indicated that the corneal mechanical properties change earlier than morphological variations in the early stages of keratoconus.4,5 Therefore, understanding the biomechanical properties of the early keratoconus is helpful for early diagnosis of keratoconus. 
The Corneal Visualization Scheimpflug Technology (Corvis ST; Oculus, Wetzlar, Germany) is currently the most commonly utilized device for evaluating corneal biomechanical properties. It applies an air puff force to the central corneal area and provides a series of dynamic corneal response (DCR) parameters to characterize corneal mechanical properties in vivo. In the early stages of keratoconus, it is generally necessary to combine corneal tomography and DCR parameters for clinical diagnosis.6 Although biomechanical combination parameters such as Corneal Biomechanical Index (CBI) and methods combining Corvis ST with corneal tomography data have achieved significant success in diagnosing early and clinical keratoconus, but with limited diagnostic efficacy for early keratoconus.712 An important aspect is that the DCR parameters are not direct mechanical parameters of the cornea, such as elastic modulus. The relationship between DCR parameters and corneal biomechanical properties remains unclear, impeding our comprehensive understanding of corneal biomechanical properties in keratoconus patients. Therefore, it is necessary to develop keratoconus models that can characterize the main morphological and mechanical features of keratoconus to study the dynamic response of the cornea under an air puff. The corresponding findings could contribute to establishing associations between DCR parameters and corneal mechanical properties. 
The DCR parameters are influenced by various factors, including corneal morphology (such as corneal curvature and thickness), intraocular pressure (IOP), elastic modulus, and age.13 Some researchers have developed animal models of corneal ectasia14 that exhibit characteristics similar to those of clinical keratoconus, including a reduction in central corneal thickness and localized softening. However, due to significant individual differences, the high cost of modeling, and numerous uncontrollable variables, relatively few studies have utilized keratoconus animal models to explore DCR parameters. 
Another different strategy is to create artificial corneal models made of silicone. Researchers have developed various experimental corneal models for different research purposes.15-18 However, all of them are normal corneal models, and no keratoconus models have been reported. Silicone, as a material with an adjustable elastic modulus and easy accessibility, can be utilized to prepare corneal models that closely resemble the morphology and elastic modulus of the human cornea. In this study, we prepared normal corneal phantoms (NCPs) using silicone, as well as early keratoconus phantoms with locally weakened mechanical properties and tested the fabricated corneal models using Corvis ST. Our goal was to provide a method for constructing early keratoconus models and a basis for studying the dynamic response of the cornea using early keratoconus phantoms. 
Materials and Methods
Corneal Mold Design
Based on human corneal dimensions,19,20 computer-aided design (CAD) software was used to draw the anterior and posterior surface contour curves of the corneal phantom (Fig. 1a). In this study, the anterior surface of the designed cornea was generated by rotating a segment of an ellipse around its minor axis. Therefore, the apex of the designed cornea was located at the endpoint of the minor axis of the ellipse. The major semi-axis of the ellipse was 8.4 mm, and the minor semi-axis was 7.5 mm. The posterior surface was generated by rotating a segment of a circle with a radius of 7 mm. Overall, the corneal model had a diameter of 12 mm, a central thickness of 0.5 mm, an apical radius curvature of 9.4 mm, and an edge thickness of 1 mm (Fig. 1a). Based on the aforementioned geometric parameters, a mold was designed (Fig. 1b) comprised of the base and the lid. During manufacturing, an appropriate amount of silicone is poured into the base, followed by the placement of the lid. Subsequently, any excess silicone flows out through the open hole on the side. 
Figure 1.
 
Design of the corneal mold. (a) The morphological design of the cornea phantom. (b) The corneal phantom mold, with the base on the left and the lid on the right.
Figure 1.
 
Design of the corneal mold. (a) The morphological design of the cornea phantom. (b) The corneal phantom mold, with the base on the left and the lid on the right.
Material Fabrication and Mechanical Tests
Using an appropriate amount of 107 Silicone (Jianhao Electronics, Tianchang, China) with a viscosity of 10,000 cp, we first added 5% crosslinking agent (tetraethyl orthosilicate) and 2% catalyst (dibutyltin dilaurate), mixing well after the addition of a specific amount of silicone oil (Sanjingxinde Technology, Beijing, China) to adjust the rigidity of the silicone after curing. We then added 5% of white carbon black to adjust the transparency of the corneal phantoms after molding. The crosslinking agent, catalyst, and white carbon black were obtained from Shanghai Macklin Biochemical (Shanghai, China). After all the ingredients were well mixed, they were placed in the vacuum device to eliminate bubbles in the silicone. 
The prepared silicone was divided into two parts: One part was placed in the corneal phantom mold and allowed to cure for 12 hours at room temperature, and the other part was cut into silicone strips after curing and was used to determine the elastic modulus of the silicone material. In this study, 16.67%, 33.33%, 44.44%, and 50% mass fractions of silicone oil (relative to the total mass of the mixture) were added to 107 Silicone to obtain four silicone materials with different elastic moduli. 
After the second part of the silicone material was cured, it was cut into strips approximately 2 mm thick and 40 mm long using a parallel double-edged knife with a width of 4 mm. Five specimens of each silicone material were prepared for uniaxial tensile test at a tensile rate of 0.02 mm/s and a tensile length of 5 mm. The width and thickness of the silicone strips were measured using a thickness gauge, and the length was measured using a caliper.21 
Fabrication of the Early Keratoconus Phantoms
Local biomechanical properties undergo change in the early stages of keratoconus, yet the geometry of the cornea remains relatively unaltered. In this study, two types of early keratoconus models were fabricated using a two-step molding method, each made of two silicone materials with different elastic moduli to achieve the desired softness in specific areas. Based on the location of the softened area, the two early keratoconus models were designated as central keratoconus phantom (CKP) and paracentral keratoconus phantom (PKP) to simulate early keratoconus. 
The higher the silicone oil content, the softer the silicone material becomes. First, a silicone sample with a high silicone oil content was prepared. Then, a small amount of this sample was dropped with a syringe needle into the target areas (the central or paracentral areas) of the mold and covered with the lid. After the silicone material was cured, an adequate amount of silicone sample with a low silicone oil content was added to the mold for the second molding. This process produced locally softened corneal phantoms that retained a normal morphology. 
To control the location of the softened area, the mold was placed on a horizontal table top and the softer silicone was dropped into the central area of the mold (Fig. 2a); thus, the softened area was located in the central area of the cornea phantom (Fig. 2b). If one end ofthe mold was elevated to a specific height (Fig. 2d), the silicone droplet gathered in the paracentral area of the mold, resulting in softening occurring in the paracentral area of the cornea (Fig. 2e). To control the size of the softened area, the syringe needle was vertically immersed in the silicone sample to a specific depth, causing the silicone to adhere to the outer surface of the needle. Gravity then caused the silicone droplet to form and drip into the mold (Figs. 2a, 2d). The volume of the silicone droplet depended on the size of the needle, the depth to which the needle was inserted, and the viscosity of the silicone. By ensuring consistency in these three factors in each experiment, a similar volume of silicone droplets was obtained, resulting in uniformity in the size of the softened area. The reliability of the method was confirmed by measuring the diameter of the softened area (Figs. 2c, 2f). 
Figure 2.
 
Preparation and measurement of the softened area of early keratoconus phantoms. (a) The silicone is dropped into the central area. (b) The silicone droplet is cured in the central area. (c) Measurement of the diameter of the central softened area. (d) The silicone is dropped into the paracentral area. (e) The silicone droplet is cured in the paracentral area. (f) Measurement of the diameter of the paracentral softened area.
Figure 2.
 
Preparation and measurement of the softened area of early keratoconus phantoms. (a) The silicone is dropped into the central area. (b) The silicone droplet is cured in the central area. (c) Measurement of the diameter of the central softened area. (d) The silicone is dropped into the paracentral area. (e) The silicone droplet is cured in the paracentral area. (f) Measurement of the diameter of the paracentral softened area.
Test Platform Setup and Data Acquisition
The Corvis ST test platform for corneal phantoms includes an artificial anterior chamber, the Corvis ST, a microinjection pump, and a pressure sensor.22 Before testing, the corneal phantoms were mounted on the artificial anterior chamber and then the artificial anterior chamber, microinjection pump, and pressure sensor were connected. The system was connected to the atmosphere to stabilize its pressure. During the test, the artificial anterior chamber was fixed directly in front of the Corvis ST (for PKPs, the center of the softened area was located on the horizontal meridian). The microinjection pump was used to inject physiological saline into the artificial anterior chamber at a rate of 50 µL/min to adjust the internal pressure of the artificial anterior chamber to simulate IOP. The simulated intraocular pressure (SIOP) was measured with the pressure sensor. In this study, the SIOP was set to seven levels within a range of 8 to 28 mmHg. When the predetermined value was reached, the microinjection pump was stopped, and the corneal phantom was tested with the Corvis ST. Three repeated measurements were taken at each pressure point to minimize random error. 
In this study, we selected 12 DCR parameters for analysis: deflection amplitude of the cornea at the highest concavity (HCDA), distance between the two peaks of the cornea at the highest concavity (PD), corneal curvature radius at highest concavity (HR), time from the start of the air puff until the cornea reaches its highest concavity (HCT), maximum corneal velocity at the first and second applanation (A1V and A2V), first and second applanation times (A1T and A2T), length of the flattened cornea at the first and second applanation (A1L and A2L), integrated radius (IR), and the stiffness parameter at the first applanation (SPA1). 
Statistical Analysis
Results are expressed as mean ± SD. The independent sample t-test was used to compare the two groups. The Spearman correlation test was used to analyze the correlation. Repeated-measures analysis of variance (ANOVA) was used to compare the overall differences among the three types of cornea phantoms (NCP, CKP, and PKP), and one-way ANOVA was used to compare their differences at each pressure point. Differences among groups were analyzed using the least significant difference (LSD) post hoc test. All statistical analyses were performed with SPSS Statistics 21.0 (IBM, Chicago, IL), and P < 0.05 was considered statistically significant. 
Results
Elastic Moduli of Silicone Materials
The test results of elastic moduli of silicone materials with different silicone oil ratios are shown in Figure 3. The stress and strain of silicone materials exhibit an essentially linear relationship within the tensile range (Fig. 3a). The elastic modulus of silicone material is almost linearly related to the content of silicone oil (Fig. 3b). The elastic moduli of silicone materials with 16.67% to 50% silicone oil ranged from 0.134 to 0.451 MPa. 
Figure 3.
 
Measurement of the elastic moduli of silicone strips. (a) Stress–strain curves for silicone strips for four different silicone oil contents. (b) Elastic moduli of silicone materials with different silicone oil ratios.
Figure 3.
 
Measurement of the elastic moduli of silicone strips. (a) Stress–strain curves for silicone strips for four different silicone oil contents. (b) Elastic moduli of silicone materials with different silicone oil ratios.
Fabrication and Characterization of Corneal Phantoms
The prepared silicone was poured into the mold and allowed to cure for 12 hours, and then the molded silicone was removed from the mold to obtain NCPs (Fig. 3a). As shown in Figure 4b, a 2-mm margin was retained on the corneal phantom to facilitate its mounting in the artificial anterior chamber for the Corvis ST test (Fig. 4c). The diameters of the softened area for the early keratoconus phantoms are shown in Table 1. A total of six cornea phantoms were tested for each type of early keratoconus phantom (CKP and PKP). The mean diameter of the softened areas was 4.56 ± 0.06 mm for CKPs and 4.59 ± 0.09 mm for PKPs. No statistically significant difference was observed between the groups. 
Figure 4.
 
Fabrication and mounting of the corneal phantom. (a) Cured silicone being removed from the mold. (b) A normal corneal phantom. (c) A normal corneal phantom mounted in the artificial anterior chamber.
Figure 4.
 
Fabrication and mounting of the corneal phantom. (a) Cured silicone being removed from the mold. (b) A normal corneal phantom. (c) A normal corneal phantom mounted in the artificial anterior chamber.
Table 1.
 
Diameters (mm) of the Softened Areas of Two Types of Early Keratoconus Phantoms
Table 1.
 
Diameters (mm) of the Softened Areas of Two Types of Early Keratoconus Phantoms
The two types of locally softened corneal phantoms prepared by the two-step molding method are shown in Figures 5a and 5c. The outline of the softened area is faintly visible. The center of the softened area in the CKP coincides with the corneal apex, but in the PKP the center of the softened area is approximately 2 mm horizontally from the corneal apex. The two locally softened corneal phantoms were mounted on the artificial anterior chamber to conduct pressure tests via water injection. As illustrated in Figures 5b and 5d, higher pressure resulted in a more pronounced degree of expansion. Nevertheless, no noticeable local protrusion or any signs of cracks or leaks were observed. 
Figure 5.
 
The locally softened corneal phantoms and their morphological changes at different water pressures. (a) CKP. (b) CKP under water pressure. (c) PKP. (d) PKP under water pressure.
Figure 5.
 
The locally softened corneal phantoms and their morphological changes at different water pressures. (a) CKP. (b) CKP under water pressure. (c) PKP. (d) PKP under water pressure.
The Oculus Pentacam can provide high-resolution images of corneal thickness distribution. To verify whether the thickness distribution of the corneal phantom fabricated by the mold was within normal limits, we used the Pentacam to measure the thickness of a normal corneal phantom (Fig. 6). The results showed that the thinnest point of the corneal phantom was located in the central region, with thickness gradually increasing from the center toward the periphery. Specifically, the central thickness of the corneal phantom was approximately 500 µm, which is consistent with the designed dimensions. 
Figure 6.
 
An example of the thickness distribution of a normal corneal phantom tested by the Pentacam.
Figure 6.
 
An example of the thickness distribution of a normal corneal phantom tested by the Pentacam.
Test Results of Normal Corneal Phantoms
Four corneal models with different elastic moduli were selected for the Corvis ST test, with five samples for each model. HCDA was selected to compare the four corneal models and normal human eyes. The results showed that the HCDA of the corneal model with 44.44% silicone oil (with a corresponding elastic modulus of 0.182 MPa) within the normal IOP range was comparable to that of normal human eyes.23 Figure 7 illustrates the time-dependent changes in corneal deflection amplitude at 15.15 mmHg for a normal corneal phantom and a normal human cornea. It can be observed that the maximum deflection amplitudes of both samples are almost identical, and the temporal evolutions of their curves exhibit considerable consistency. 
Figure 7.
 
Deflection amplitude with temporal evolution of a normal human cornea and a normal corneal phantom.
Figure 7.
 
Deflection amplitude with temporal evolution of a normal human cornea and a normal corneal phantom.
In order to investigate the variation of DCR parameters with SIOP, the normal corneal phantoms with an elastic modulus of 0.182 MPa were selected to simulate normal human cornea for Corvis ST testing. The results are shown in Figure 8. Correlation analysis showed that HCDA, PD, HCT, A1V, A2T, and IR were negatively correlated with SIOP, whereas HR, A2V, A1T, and SPA1 were positively correlated with SIOP. Meanwhile, A1L and A2L showed no significant correlation with SIOP. 
Figure 8.
 
Variation trend of DCR parameters with SIOP in NCPs (elastic modulus, 0.182 MPa). (a–l) Variations in HCDA, HR, PD, HCT, A1V, A2V, A1T, A2T, A1L, A2L, IR, and SPA1 with SIOP, where r is the Spearman correlation coefficient between DCR parameters and SIOP.
Figure 8.
 
Variation trend of DCR parameters with SIOP in NCPs (elastic modulus, 0.182 MPa). (a–l) Variations in HCDA, HR, PD, HCT, A1V, A2V, A1T, A2T, A1L, A2L, IR, and SPA1 with SIOP, where r is the Spearman correlation coefficient between DCR parameters and SIOP.
Test Results of Early Keratoconus Phantoms
The modulus of the softened area in the two early keratoconus models was set at 0.134 MPa, and the modulus of the other region was maintained at 0.182 MPa, consistent with that of the normal corneal model. Corvis ST testing was performed on three groups of corneal models, including NCPs and the two types of early keratoconus phantoms (CKP and PKP). Each group consisted of five samples, and DCR parameters were obtained for comparison. Table 2 shows the P values of several DCR parameters for pairwise comparisons of the three groups of corneal phantoms. Significant differences were found in HCDA, PD, HR, A1V, A2V, A1T, A2T, and IR between the early keratoconus phantoms and NCP, as well as significant differences in HCDA, HR, A1V, A1T, and IR between the two early keratoconus phantoms. The differences in DCR parameters, which showed overall differences among the three groups of corneal phantoms, were also compared under each specific SIOP (Fig. 9). The trend of DCR parameters with SIOP for the two types of early keratoconus phantoms was basically consistent with the results obtained for NCPs (Fig. 9). 
Table 2.
 
DCR parameters P Values for Pairwise Comparisons of the Three Groups of Corneal Phantoms
Table 2.
 
DCR parameters P Values for Pairwise Comparisons of the Three Groups of Corneal Phantoms
Figure 9.
 
Variation trend of DCR parameters with SIOP in normal corneal phantoms and two types of early keratoconus phantoms. (al) Variations in HCDA, HR, PD, HCT, A1V, A2V, A1T, A2T, A1L, A2L, IR, and SPA1 with SIOP. One-way ANOVA with the LSD post hoc test was used to compare their differences at each pressure point. CKP versus NCP: *P < 0.05, **P < 0.01; PKP versus NCP: #P < 0.05, ##P < 0.01; CKP versus PKP: +P < 0.05, ++P < 0.01.
Figure 9.
 
Variation trend of DCR parameters with SIOP in normal corneal phantoms and two types of early keratoconus phantoms. (al) Variations in HCDA, HR, PD, HCT, A1V, A2V, A1T, A2T, A1L, A2L, IR, and SPA1 with SIOP. One-way ANOVA with the LSD post hoc test was used to compare their differences at each pressure point. CKP versus NCP: *P < 0.05, **P < 0.01; PKP versus NCP: #P < 0.05, ##P < 0.01; CKP versus PKP: +P < 0.05, ++P < 0.01.
Discussion
In this study, we developed a method to construct normal corneal and early keratoconus phantom models. The NCP model was fabricated using a self-designed mold. Based on this mold, two types of early keratoconus phantoms with locally softened areas were fabricated using a two-step molding process. The results showed that the overall morphology and thickness distribution of the corneal phantoms fabricated by our mold were basically consistent with the designed size and could be used for the Corvis ST test. The DCR parameters HCDA, PD, HR, A1V, A2V, A1T, A2T, IR, and SPA1 of the corneal phantoms were significantly correlated with SIOP. HCDA, PD, HR, A1V, A2V, A1T, A2T, and IR of early keratoconus phantoms were significantly different from those of the NCPs. These results suggest that our method for generating the corneal phantoms is both feasible and effective, and the corneal phantoms generated using this method are well suited for investigating the mechanical behavior of the cornea under an air puff; the pertinent findings will facilitate a more profound comprehension of the biomechanical characteristics of keratoconus, thus offering valuable insights for the early diagnosis of keratoconus. 
Elastic modulus is an extremely important parameter that characterizes the mechanical properties of the cornea. Based on data from the Corvis ST test, the range of human corneal elastic modulus is 0.05 to 1.24 MPa.22,24,25 In this study, the elastic modulus of silicone materials prepared by adjusting the silicone oil content ranged from 0.13 MPa to 0.45 MPa, which is consistent with human corneal elastic modulus ranges reported in relevant literature. By comparing the HCDA of normal human corneas with those of normal corneal phantoms with four different elastic moduli, we determined that the normal corneal phantom with 44.44% silicone oil content (elastic modulus, 0.18 MPa) was suitable for simulating the normal human cornea and performing the Corvis ST test. The results showed that HCDA, PD, HCT, A1V, A2T, and IR were negatively correlated with SIOP, whereas HR, A2V, A1T, and SPA1 were positively correlated with SIOP. This result was consistent with relevant research results based on ex vivo animal eyes22,26 and the clinical test results of Wu et al.27 based on the Corvis ST. The above results indicate that the corneal phantoms can simulate the mechanical behavior of the cornea under an air puff. 
The main innovation of this study is the preparation of two types of early keratoconus phantoms using the two-step molding method. The geometry of the model matched that of the normal cornea, and the locally softened area was used to simulate the keratoconus cone area. The size of the softened area directly affects the overall mechanical response of the cornea. Therefore, it is crucial to carefully design and control the softened area. In this study, 12 early keratoconus phantoms were prepared using the two-step molding method, and the size of the softened area exhibited minimal variation (Table 1), indicating that this method could ensure consistent sizing of the softened area across different samples. The diameter of the softened area was approximately 4.5 mm, which was determined based on the size of the keratoconus cone area.28,29 The elastic modulus of the softened area was 0.134 MPa, which was 26.3% lower than that of the normal area. Zhao et al.30 determined that the difference in the elastic modulus between the keratoconus cone and its surrounding area can be up to 29%. Giraudet et al.31 established a multiscale mechanical finite element model and found that, when the local fiber stiffness of the corneal model was reduced by 30% to 40%, it was able to reproduce the keratoconus changes in SimK (a clinical indicator of corneal curvature). Therefore, the size and degree of softened area of the early keratoconus phantom designed in this study are reasonable. In addition, the early keratoconus phantoms were tested under water injection. No noticeable local protrusion was found when the pressure reached 50 mmHg, and no cracks or leaks were observed (Fig. 5), indicating that the mechanical strength of the corneal phantom prepared by the two-step molding method was acceptable.32 
The cone location is an important feature of keratoconus, and keratoconus can be classified into central keratoconus and paracentral keratoconus according to the cone location.28,33 In this paper, two types of early keratoconus phantoms (CKP and PKP) were fabricated to simulate central keratoconus and paracentral keratoconus, respectively. The Corvis ST test results showed that several DCR parameters (HCDA, HR, A1V, A1T, and IR) exhibited greater differences between CKPs and NCPs compared to the differences between PKPs and NCPs (Table 2Fig. 9). This suggests that the closer the softened area is to the geometric center of the cornea, the greater the effect on the overall mechanical properties of the cornea, resulting in softer behavior under mechanical stress. Several studies have investigated differences in the locations of various keratoconus cones, including optical differences,34 corneal topographic differences,35 differences in surgical outcomes such as corneal collagen cross-linking,36,37 and differences in the symmetry of corneal deformation.38 A recent study found that the location of the keratoconus cone affects corneal biomechanical parameters measured by the Ocular Response Analyzer (Reichert Technologies, Depew, NY).39 The paracentral keratoconus shows greater stiffness than the central keratoconus, which is consistent with our results. Such findings indicate that the paracentral keratoconus is less likely to be detected than the central keratoconus, which may explain why some patients who undergo refractive surgery are at risk of developing corneal ectasia after the surgery.40 
DCR parameters can be used to distinguish clinical keratoconus from normal corneas,41-45 but there is no consensus on the ability of DCR parameters to distinguish early keratoconus. One reason is that the response of DCR parameters to changes in the local mechanical properties of the cornea is not fully understood. Our results show that the HCDA, PD, HR, A1V, A1T, A2T, and IR of normal and early keratoconus phantoms are significantly different in the normal range of IOP, so these parameters and their changes should receive more attention in the diagnosis of keratoconus. Recently, artificial intelligence methods have shown significant application potential in diagnosing keratoconus; however, further algorithmic optimization is required for the early diagnosis of keratoconus.8,46 Introducing the influence of changes in corneal mechanical properties on DCR parameters as a prior knowledge into these algorithms may improve their diagnostic performance.47,48 
The distribution of collagen fiber bundles in normal corneas exhibits directionality and density variations, which lead to anisotropic mechanical properties in the cornea.49,50 This study ignored the anisotropic mechanical properties of the cornea and used isotropic silicone to construct corneal phantoms for investigating the corneal mechanical response under an air puff. The results show that the DCR parameters of the normal corneal model exhibit a response trend to SIOP changes similar to that observed in ex vivo animal eyes. The mechanical properties of keratoconus exhibit regional inhomogeneity. Compared to the anisotropic mechanical properties of the cornea, the regional inhomogeneity of the mechanical properties of keratoconus has a more significant impact on DCR parameters. Based on these considerations, this study ignored the anisotropic mechanical properties of the cornea when constructing normal corneal phantoms and early keratoconus corneal phantoms. This approach allowed us to eliminate additional variables, ensuring that the experimental results would directly reflect the macroscopic effects of the mechanical property changes. 
In conclusion, we proposed a method for fabricating early keratoconus phantom models by using two silicone materials with different elastic moduli to modify the local mechanical properties of the corneal models. These corneal models can be used to study the biomechanical behavior of keratoconus under an air puff. In the future, we can utilize this method to create keratoconus models with different cone characteristics (such as cone size, cone position, and cone stiffness), to study the effects of these characteristics on DCR parameters, and to combine them with clinical data to enhance the performance of the Corvis ST in diagnosing early keratoconus. 
Acknowledgments
Supported by a grant from the National Natural Science Foundation of China (32171304). 
Disclosure: H. Tong, None; M. Wu, None; J. Han, None; L. Li, None; H. Zhang, None 
References
Randleman JB, Russell B, Ward MA, Thompson KP, Stulting RD. Risk factors and prognosis for corneal ectasia after LASIK. Ophthalmology. 2003; 110(2): 267–275. [CrossRef] [PubMed]
Gomes JAP, Tan D, Rapuano CJ, et al. Global consensus on keratoconus and ectatic diseases. Cornea. 2015; 34(4): 359–369. [CrossRef] [PubMed]
Henriquez MA, Hadid M, Izquierdo L. A systematic review of subclinical keratoconus and forme fruste keratoconus. J Refract Surg. 2020; 36(4): 270–279. [CrossRef] [PubMed]
Ruberti JW, Sinha Roy A, Roberts CJ. Corneal biomechanics and biomaterials. Annu Rev Biomed Eng. 2011; 13(1): 269–295. [CrossRef] [PubMed]
Ali NQ, Patel DV, McGhee CNJ. Biomechanical responses of healthy and keratoconic corneas measured using a noncontact Scheimpflug-based tonometer. Invest Ophthalmol Vis Sci. 2014; 55(6): 3651–3659. [CrossRef] [PubMed]
Esporcatte LPG, Salomão MQ, Lopes BT, et al. Biomechanics in keratoconus diagnosis. Curr Eye Res. 2023; 48(2): 130–136. [CrossRef] [PubMed]
Steinberg J, Siebert M, Katz T, et al. Tomographic and biomechanical Scheimpflug imaging for keratoconus characterization: a validation of current indices. J Refract Surg. 2018; 34(12): 840–847. [CrossRef] [PubMed]
Ambrósio R, Machado AP, Leão E, et al. Optimized artificial intelligence for enhanced ectasia detection using Scheimpflug-based corneal tomography and biomechanical data. Am J Ophthalmol. 2023; 251: 126–142. [CrossRef] [PubMed]
Miao YY, Ma XM, Qu ZX, et al. Performance of Corvis ST parameters including updated stress-strain index in differentiating between normal, forme-fruste, subclinical, and clinical keratoconic eyes. Am J Ophthalmol. 2024; 258: 196–207. [CrossRef] [PubMed]
Ambrósio R, Lopes BT, Faria-Correia F, et al. Integration of Scheimpflug-based corneal tomography and biomechanical assessments for enhancing ectasia detection. J Refract Surg. 2017; 33(7): 434–443. [CrossRef] [PubMed]
Lu N-J, Koppen C, Hafezi F, et al. Combinations of Scheimpflug tomography, ocular coherence tomography and air-puff tonometry improve the detection of keratoconus. Cont Lens Anterior Eye. 2023; 46(3): 101840. [CrossRef] [PubMed]
Tian L, Zhang D, Guo L, et al. Comparisons of corneal biomechanical and tomographic parameters among thin normal cornea, forme fruste keratoconus, and mild keratoconus. Eye Vis (Lond). 2021; 8(1): 1–11. [PubMed]
Fontes BM, Ambrósio R, Jr, Velarde GC, Nosé W. Corneal biomechanical evaluation in healthy thin corneas compared with matched keratoconus cases. Arq Bras Oftalmol. 2011; 74: 13–16. [CrossRef] [PubMed]
Wei J, He R, Wang X, et al. The corneal ectasia model of rabbit: a validity and stability study. Bioengineering (Basel). 2023; 10(4): 479. [CrossRef] [PubMed]
Wang L, Tian L, Huang Y, Huang Y, Zheng Y. Assessment of corneal biomechanical properties with inflation test using optical coherence tomography. Ann Biomed Eng. 2018; 46(2): 247–256. [CrossRef] [PubMed]
Ma G, Cai J, Zhong R, et al. Corneal surface wave propagation associated with intraocular pressures: oct elastography assessment in a simplified eye model. Bioengineering (Basel). 2023; 10(7): 754. [CrossRef] [PubMed]
Glass DH, Roberts CJ, Litsky AS, Weber PA. A viscoelastic biomechanical model of the cornea describing the effect of viscosity and elasticity on hysteresis. Invest Ophthalmol Vis Sci. 2008; 49(9): 3919–3926. [CrossRef] [PubMed]
Cho HS, Jeoung SC, Yang YS. Development of eye phantom for mimicking the deformation of the human cornea accompanied by intraocular pressure alterations. Sci Rep. 2022; 12(1): 20670. [CrossRef] [PubMed]
Burek H, Douthwaite WA. Mathematical models of the general corneal surface. Ophthalmic Physiol Opt. 1993; 13(1): 68–72. [CrossRef] [PubMed]
Su P, Yang Y, Song Y. Corneal hyper-viscoelastic model: derivations, experiments, and simulations. Acta Bioeng Biomech. 2015; 17(2): 73–84. [PubMed]
Zhang D, Qin X, Zhang H, Li L. Time-varying regularity of changes in biomechanical properties of the corneas after removal of anterior corneal tissue. Biomed Eng Online. 2021; 20(1): 113. [CrossRef] [PubMed]
Qin X, Tian L, Zhang H, Chen X, Li L. Evaluation of corneal elastic modulus based on corneal visualization Scheimpflug technology. Biomed Eng Online. 2019; 18: 42. [CrossRef] [PubMed]
Borderie V, Beauruel J, Cuyaubère R, Georgeon C, Memmi B, Sandali O. Comprehensive assessment of Corvis ST biomechanical indices in normal and keratoconus corneas with reference to corneal enantiomorphism. J Clin Med. 2023; 12(2): 690. [CrossRef] [PubMed]
Shih PJ, Huang CJ, Huang TH, et al. Estimation of the corneal Young's modulus in vivo based on a fluid-filled spherical-shell model with Scheimpflug imaging. J Ophthalmol. 2017; 2017(1): 5410143. [PubMed]
Shih PJ, Cao HJ, Huang CJ, Wang IJ, Shih WP, Yen JY. A corneal elastic dynamic model derived from Scheimpflug imaging technology. Ophthalmic Physiol Opt. 2015; 35(6): 663–672. [CrossRef] [PubMed]
Bao F, Deng M, Wang Q, et al. Evaluation of the relationship of corneal biomechanical metrics with physical intraocular pressure and central corneal thickness in ex vivo rabbit eye globes. Exp Eye Res. 2015; 137: 11–17. [CrossRef] [PubMed]
Wu Y, Tian L, Fei HY. In vivo corneal biomechanical properties with corneal visualization Scheimpflug technology in Chinese population. Biomed Res Int. 2016; 2016(1): 7840284. [PubMed]
Romero-Jiménez M, Santodomingo-Rubido J, Wolffsohn JS. Keratoconus: a review. Contact Lens Anterior Eye. 2010; 33(4): 157–166. [CrossRef] [PubMed]
Santodomingo-Rubido J, Carracedo G, Suzaki A, Villa-Collar C, Vincent SJ, Wolffsohn JS. Keratoconus: an updated review. Contact Lens Anterior Eye. 2022; 45(3): 101559. [CrossRef] [PubMed]
Zhao Y, Yang H, Li Y, et al. Quantitative assessment of biomechanical properties of the human keratoconus cornea using acoustic radiation force optical coherence elastography. Transl Vis Sci Technol. 2022; 11(6): 4. [CrossRef]
Giraudet C, Diaz J, Le Tallec P, Allain JM. Multiscale mechanical model based on patient-specific geometry: application to early keratoconus development. J Mech Behav Biomed Mater. 2022; 129: 105121. [CrossRef] [PubMed]
Andersson J, Gubanski SM, Hillborg H. Properties of interfaces in silicone rubber. IEEE Trans Dielectr Electr Insul. 2007; 14(1): 137–145. [CrossRef]
Rabinowitz YS . Keratoconus. Surv Ophthalmol. 1998; 42(4): 297–319. [CrossRef] [PubMed]
Tan B, Baker K, Chen YL, et al. How keratoconus influences optical performance of the eye. J Vis. 2008; 8(2): 13. [CrossRef]
Prakash G, Srivastava D, Choudhuri S, Thirumalai SM, Bacero R. Differences in central and non-central keratoconus, and their effect on the objective screening thresholds for keratoconus. Acta Ophthalmol. 2016; 94(2): e118–e129. [CrossRef] [PubMed]
Greenstein SA, Fry KL, Hersh PS. Effect of topographic cone location on outcomes of corneal collagen cross-linking for keratoconus and corneal ectasia. J Refract Surg. 2012; 28(6): 397–405. [CrossRef] [PubMed]
Shetty R, Nuijts RMMA, Nicholson M, et al. Cone location–dependent outcomes after combined topography-guided photorefractive keratectomy and collagen cross-linking. Am J Ophthalmol. 2015; 159(3): 419–425. [CrossRef] [PubMed]
Curatolo A, Birkenfeld JS, Martinez-Enriquez E, et al. Multi-meridian corneal imaging of air-puff induced deformation for improved detection of biomechanical abnormalities. Biomed Opt Express. 2020; 11(11): 6337–6355. [CrossRef] [PubMed]
Yuhas PT, Fortman MM, Mahmoud AM, Roberts CJ. Keratoconus cone location influences ocular biomechanical parameters measured by the Ocular Response Analyzer. Eye Vis (Lond). 2024; 11(1): 2. [CrossRef] [PubMed]
Moshirfar M, Tukan AN, Bundogji N, et al. Ectasia after corneal refractive surgery: a systematic review. Ophthalmol Ther. 2021; 10(4): 753–776. [CrossRef] [PubMed]
Tian L, Qin X, Zhang H, et al. A potential screening index of corneal biomechanics in healthy subjects, forme fruste keratoconus patients and clinical keratoconus patients. Front Bioeng Biotechnol. 2021; 9: 76605.
Karimi A, Meimani N, Razaghi R, Rahmati SM, Jadidi K, Rostami M. Biomechanics of the healthy and keratoconic corneas: a combination of the clinical data, finite element analysis, and artificial neural network. Curr Pharm Des. 2018; 24(37): 4474–4483. [CrossRef] [PubMed]
Elham R, Jafarzadehpur E, Hashemi H, et al. Comparisons of corneal biomechanical and tomographic parameters among thin normal cornea, forme fruste keratoconus, and mild keratoconus. J Curr Ophthalmol. 2017; 29(3): 175–181. [CrossRef] [PubMed]
Catalán-López S, Cadarso-Suárez L, López-Ratón M, Cadarso-Suárez C. Corneal biomechanics in unilateral keratoconus and fellow eyes with a Scheimpflug-based tonometer. Optom Vis Sci. 2018; 95(7): 608–615. [CrossRef] [PubMed]
Steinberg J, Katz T, Lücke K, Frings A, Druchkiv V, Linke SJ. Screening for keratoconus with new dynamic biomechanical in vivo Scheimpflug analyses. Cornea. 2015; 34(11): 1404–1412. [CrossRef] [PubMed]
Huo Y, Chen X, Khan GA, Wang Y. Corneal biomechanics in early diagnosis of keratoconus using artificial intelligence. Clin Exp Ophthalmol. 2024; 262(4): 1337–1349.
Von Rueden L, Mayer S, Beckh K, et al. Informed machine learning–a taxonomy and survey of integrating prior knowledge into learning systems. IEEE Trans Knowl Data Eng. 2023; 35(1): 614–633.
Roychowdhury S, Diligenti M, Gori M. Regularizing deep networks with prior knowledge: a constraint-based approach. Knowl-Based Syst. 2021; 222: 106989. [CrossRef]
Winkler M, Chai D, Kriling S, et al. Nonlinear optical macroscopic assessment of 3-D corneal collagen organization and axial biomechanics. Invest Ophthalmol Vis Sci. 2011; 52(12): 8818–8827. [CrossRef] [PubMed]
Winkler M, Shoa G, Xie Y, et al. Three-dimensional distribution of transverse collagen fibers in the anterior human corneal stroma. Invest Ophthalmol Vis Sci. 2013; 54(12): 7293–7301. [CrossRef] [PubMed]
Figure 1.
 
Design of the corneal mold. (a) The morphological design of the cornea phantom. (b) The corneal phantom mold, with the base on the left and the lid on the right.
Figure 1.
 
Design of the corneal mold. (a) The morphological design of the cornea phantom. (b) The corneal phantom mold, with the base on the left and the lid on the right.
Figure 2.
 
Preparation and measurement of the softened area of early keratoconus phantoms. (a) The silicone is dropped into the central area. (b) The silicone droplet is cured in the central area. (c) Measurement of the diameter of the central softened area. (d) The silicone is dropped into the paracentral area. (e) The silicone droplet is cured in the paracentral area. (f) Measurement of the diameter of the paracentral softened area.
Figure 2.
 
Preparation and measurement of the softened area of early keratoconus phantoms. (a) The silicone is dropped into the central area. (b) The silicone droplet is cured in the central area. (c) Measurement of the diameter of the central softened area. (d) The silicone is dropped into the paracentral area. (e) The silicone droplet is cured in the paracentral area. (f) Measurement of the diameter of the paracentral softened area.
Figure 3.
 
Measurement of the elastic moduli of silicone strips. (a) Stress–strain curves for silicone strips for four different silicone oil contents. (b) Elastic moduli of silicone materials with different silicone oil ratios.
Figure 3.
 
Measurement of the elastic moduli of silicone strips. (a) Stress–strain curves for silicone strips for four different silicone oil contents. (b) Elastic moduli of silicone materials with different silicone oil ratios.
Figure 4.
 
Fabrication and mounting of the corneal phantom. (a) Cured silicone being removed from the mold. (b) A normal corneal phantom. (c) A normal corneal phantom mounted in the artificial anterior chamber.
Figure 4.
 
Fabrication and mounting of the corneal phantom. (a) Cured silicone being removed from the mold. (b) A normal corneal phantom. (c) A normal corneal phantom mounted in the artificial anterior chamber.
Figure 5.
 
The locally softened corneal phantoms and their morphological changes at different water pressures. (a) CKP. (b) CKP under water pressure. (c) PKP. (d) PKP under water pressure.
Figure 5.
 
The locally softened corneal phantoms and their morphological changes at different water pressures. (a) CKP. (b) CKP under water pressure. (c) PKP. (d) PKP under water pressure.
Figure 6.
 
An example of the thickness distribution of a normal corneal phantom tested by the Pentacam.
Figure 6.
 
An example of the thickness distribution of a normal corneal phantom tested by the Pentacam.
Figure 7.
 
Deflection amplitude with temporal evolution of a normal human cornea and a normal corneal phantom.
Figure 7.
 
Deflection amplitude with temporal evolution of a normal human cornea and a normal corneal phantom.
Figure 8.
 
Variation trend of DCR parameters with SIOP in NCPs (elastic modulus, 0.182 MPa). (a–l) Variations in HCDA, HR, PD, HCT, A1V, A2V, A1T, A2T, A1L, A2L, IR, and SPA1 with SIOP, where r is the Spearman correlation coefficient between DCR parameters and SIOP.
Figure 8.
 
Variation trend of DCR parameters with SIOP in NCPs (elastic modulus, 0.182 MPa). (a–l) Variations in HCDA, HR, PD, HCT, A1V, A2V, A1T, A2T, A1L, A2L, IR, and SPA1 with SIOP, where r is the Spearman correlation coefficient between DCR parameters and SIOP.
Figure 9.
 
Variation trend of DCR parameters with SIOP in normal corneal phantoms and two types of early keratoconus phantoms. (al) Variations in HCDA, HR, PD, HCT, A1V, A2V, A1T, A2T, A1L, A2L, IR, and SPA1 with SIOP. One-way ANOVA with the LSD post hoc test was used to compare their differences at each pressure point. CKP versus NCP: *P < 0.05, **P < 0.01; PKP versus NCP: #P < 0.05, ##P < 0.01; CKP versus PKP: +P < 0.05, ++P < 0.01.
Figure 9.
 
Variation trend of DCR parameters with SIOP in normal corneal phantoms and two types of early keratoconus phantoms. (al) Variations in HCDA, HR, PD, HCT, A1V, A2V, A1T, A2T, A1L, A2L, IR, and SPA1 with SIOP. One-way ANOVA with the LSD post hoc test was used to compare their differences at each pressure point. CKP versus NCP: *P < 0.05, **P < 0.01; PKP versus NCP: #P < 0.05, ##P < 0.01; CKP versus PKP: +P < 0.05, ++P < 0.01.
Table 1.
 
Diameters (mm) of the Softened Areas of Two Types of Early Keratoconus Phantoms
Table 1.
 
Diameters (mm) of the Softened Areas of Two Types of Early Keratoconus Phantoms
Table 2.
 
DCR parameters P Values for Pairwise Comparisons of the Three Groups of Corneal Phantoms
Table 2.
 
DCR parameters P Values for Pairwise Comparisons of the Three Groups of Corneal Phantoms
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×