Purchase this article with an account.
Abhijit Sinha Roy, Mathew Francis, Rohit Shetty; Can corneal viscoelasticity be determined from in vivo air-puff applanation?. Invest. Ophthalmol. Vis. Sci. 2019;60(9):6819.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To assess whether human corneal viscoelasticity (CVS) can be quantified from air-puff applanation
Corvis-ST (Oculus Optikgeräte GmbH, Germany) was performed on 300 normal (N), 102 fellow eyes (FE) of KC patients, 164 grade 1 (G1), 71 grade 2 (G2) and 58 grade 3 (G3) KC eyes. Corneal and extra-corneal deformation were modeled using a 2-compartment Kelvin-Voigt (KV) setup. Calculated material parameters (MP) were corneal stiffness (kc), corneal viscosity (µc), extra-corneal stiffness (kg) and extra-corneal viscosity (µg). To assess if MP were sensitive to CVS, cornea deformation was artificially phase shifted (delayed in time) by 0.23, 0.69 and 1.16 ms (a time phase shift of 1, 3 and 5 frames, respectively). The equations were solved with non-linear least squares method. Median of MP were assessed with Kruskal-Wallis test.
In N eyes, kc was 105.5, 109.5,118.9 and 126.6 N/m for in vivo, 1, 3 and 5 frame shifts, respectively (p<0.05). In N eyes, µc was 0.0, 0.016, 0.056 and 0.099 Pa.s, respectively (p<0.05). Thus, KC model was sensitive to CVS effects, when air-puff force curve was phase shifted. Figure 1a shows comparison of in vivo and phase sifted air-puff force curves. Figure 1b shows KV model fit to the in vivo and phase shifted force curves, respectively. Here, KV was accurate in modeling all the force curves. In disease eyes, kc was 98.3, 97.2,91.0 and 81.3 N/m for FE, G1, G2 and G3 eyes, respectively (in vivo condition) [p<0.05]. In disease eyes, µc was zero for all groups (in vivo) [p>0.05]. With phase shift of 5 frames in disease eyes, kc was 118.7, 115.7, 109.9 and 97.6 N/m for FE, G1, G2 and G3 eyes, respectively (p<0.05). Similarly in disease eyes, µc was 0.088, 0.083, 0.077 and 0.061 Pa.s for FE, G1, G2 and G3 eyes, respectively (p<0.05). Thus, KV was sensitive to CVS effects in KC eyes as well, provided the force curves were phase shifted. Similar trends were observed with phase shift of 1 and 3 frames.
KV model accurately detected CVS effects in normal and KC eyes, only when the air-puff force curves were artificially phase shifted. However, CVS effects were non-existent under in vivo condition. Under in vivo condition, phase difference between force and corneal deformation was virtually non-existent. Thus, detection of CVS effects with in vivo air-puff applanation weren’t possible.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.
This PDF is available to Subscribers Only