In silico mechanical characterization of a new corneal implant for the treatment of keratoconus

Keratoconus is a progressive degenerative disease that causes the cornea to thin and protrude due to the weakening of the collagen fibers of which it is composed. This could give rise to a significant vision deterioration, potentially leading to blindness if left untreated, and a poor quality of life. Thus the need for new innovative devices to treat keratoconus is arising. The GROSSO® implant is a patented shape memory, Nickel-Titanium (Nitinol) corneal implant to reshape deformed corneas in keratoconus patients. In order to study the mechanical behaviour of the device, ideal 1D and 3D models were developed from 2D designs provided by Recornea (Rhinoceros v.7.0 (Robert McNeel & Associates, Seattle, WA, USA)). The geometric models were then meshed using beam elements and tetrahedral elements, for 1D and 3D models respectively (Hypermesh 2021 (Altair Engineering, USA)). To verify the compatibility between the ideal 3D model and the device itself, a "real” 3D model was obtained from tomographic acquisitions of real samples and meshed with the same element type as the ideal one. Then, comparisons between ideal 3D geometry and real one after electropolishing were performed through CloudCompare (GPL software, v. 2.12 beta, 2022) to assess the effect of the manufacturing process, as electropolishing. Moreover, by comparing ideal and real geometries, a geometric deviation resulting from 20 bending cycles was observed. Using finite element (FE) analyses (ABAQUS Standard (Dassault Systemes Simulia Corp., Providence, RI, USA)), displacement-driven bending simulations, and device crushing between two rigid plates were performed for all the generated models. Afterward, a mesh sensitivity analysis, gradually reducing the size of mesh elements, was carried out. Data of maximum principal strain and reaction forces were then extracted; the maximum principal strain values were compared with a safety coefficient obtained from a data sheet of Nitinol (at 22.3 °C) provided by Recornea to verify the low risk of permanent deformations during surgical implantation. The simulations showed that the maximum principal strain values of 1D, 3D, and real models (6.309%, 5.704%, and 6.659%, respectively) never reached the permanent deformation range, as desired. Regarding the maximum crushing force exerted by the rigid plates on the device, the following values were obtained for each model: 0.279 N, 0.539 N, and 0.692 N (1D, 3D, and real model, respectively). A key aspect that should be considered is that the results obtained in the bending simulation, using different model types, are comparable to each other, also proving a computational advantage provided by the use of 1D elements in FE analyses. The study can be furtherly extended by simulating risky conditions for individuals suffering from keratoconus such as eye rubbing and different types of impulsive loads or impacts on the eye.