Deriving information about biomechanical properties from highspeed-corneal imaging

Sven Reisdorf, Product Manager at OCULUS

The basic idea of the project device ImTopScanner is to derive information about the biomechanical properties from the OCT-based high-speed videos that monitor the  corneal response  to an air-pulse. Compared to the commercially already existing device based on Scheimpflug technology (Corvis ST) the OCT technology has the advantage to monitor the biomechanical response in different sectional planes. 

The following image illustrates the deformation response to the air pulse: At the beginning the cornea is in its initial convex shape. The air pulse drives the cornea backwards until the first applanation occurs when the cornea is flat and intraocular pressure (IOP) is determined. Afterwards, the cornea is further deformed until the moment of maximal concavity and then returns back to its original shape. Before it reaches the initial state, it passes through a second applanation, where the cornea is flat again. 

Figure 1: Deformation Amplitude/ Deflection Amplitude/ Whole Eye Movement

The images illustrate that one needs to distinguish between corneal movement and the movement of the whole eye as the whole globe moves slightly backwards (see image A). The red curve in the diagram illustrates the vertical displacement of the corneal vertex which is the sum of whole globe movement and corneal movement. By subtracting the movement of the whole globe from this curve one obtains the pure corneal movement (blue curve) which is the basis for further biomechanical analysis. 

Unfortunately, deriving biomechanical properties from the high-speed videos and these curves is not trivial as the movement of the cornea is not only influenced by the biomechanical properties but mainly by three factors: IOP, corneal thickness and biomechanical properties.

A lot of knowledge in regards of the influence of these main factors on the deformation response has been gained by the Corvis ST and can now be applied to OCT technology. As corneal thickness can be measured easily the main task is to differentiate the influence of IOP and biomechanical properties. In order to achieve this goal, one needs to derive features from the images that are more dependent on IOP and features that are rather influenced by the biomechanical properties. These parameters can be used clinically directly for clinical applications such as keratoconus detection and as input parameters for machine learning(1). Even more important is that they can be used as input parameters for numerical simulations that can then derive very accurately the intrinsic material properties relatively independent on corneal thickness and IOP(3). 

Dynamic Corneal Response Parameters 
As a general rule of thumb, parameters that describe the shape of the deformation are more influenced by corneal properties, whereas parameters that describe the vertical displacement of the cornea to the air pulse are more influenced by IOP. Two parameters that have been shown to be strongly related to biomechanical properties are DAratio2mm and integrated Radius: 

• DAratio2mm is a parameter that allows us to differentiate softer corneas, where the DA ratio is higher, from stiffer corneas based on the shape of deformation. The parameter has been shown to be very effective in distinguishing healthy eyes from keratoconus eyes. 

Figure 2: DADratio2mm is calculated as the ratio of deflection of the apex and 2mm nasal / temporal

• ​Another parameter that analyses the shape during the deformation is the radius of curvature that can be measured during the concave phase of the deformation. A softer tissue exhibits a smaller radius as stiffer corneas do. On the other hand, the area called integrated radius is a very good parameter to quantify the effect of corneal cross-linking. If this parameter gets smaller it indicates a stiffening of the cornea. 

Figure 3: Radius of Curvature, Inverse Concave Radius during deformation response

Structural and Material Stiffness 
Once parameters that are more influenced by biomechanical properties and less by IOP are derived, these two main factors can be further differentiated by inverse FEM simulations. Based on these simulations and analytical models, the geometrical bending stiffness can be simply determined by measuring the force that is needed to deform the cornea by a certain amount.  

Related to the ImTopScanner this bending stiffness could then be determined based on the force on the corneal surface induced by the air pulse, the intraocular pressure and the vertical displacement of the cornea. The Stiffness Parameter has been shown to be highly sensitive in regards of keratoconus detection(1,2).  

Material stiffness reflects only the intrinsic material properties determined by the structure of the tissue and not influenced by thickness or shape. Numerical simulations are required to obtain material stiffness from the high-speed videos of an OCT or Scheimpflug device. This has been achieved in cooperation with the Biomedical Engineering at University of Liverpool  by deriving a stress-strain index that represents the whole stress-strain behaviour of the cornea. This curves have been shown to be mainly independent from IOP and corneal thickness as expected(3). 

Corneal crosslinking is a key treatment for keratoconus, where mainly the material stiffness increases by creating additional crosslinks within the tissue.  When applying this concept to OCT technology, this method has the potential to optimize corneal-crosslinking procedures in order to maximize the stiffening effect of the cornea. 

Figure 5: Comparison between deformation response before CXL (red) and after CXL (green)

​(1) Vinciguerra R et al.: “Detection of Keratoconus with a New Biomechanical Index.” Journal of Refractive Surgery 32 (12), 2016: 803-810 
(2)Roberts et al.: “Introduction of Two Novel Stiffness Parameters and Interpretation of Air Puff Induced Biomechanical Deformation Response Parameters with a Dynamic Scheimpflug Analyzer.” Journal of Refractive Surgery 33(4), 2017: 266-273  
(3)Eliasy A, Chen KJ, Vinciguerra R, Lopes BT, Abass A, Vinciguerra P, Ambrósio R Jr, Roberts CJ, Elsheikh A. Determination of Corneal Biomechanical Behavior in-vivo for Healthy Eyes Using CorVis ST Tonometry: Stress-Strain Index. Front Bioeng Biotechnol. 2019;7:105.