Assessment of meibomian gland drop-out and visibility through a new quantitative method in scleral lens wearers: A one-year follow-up study

Abstract

To validate a previously developed algorithm based on the visibility of meibomian gland images obtained with Cobra fundus camera and to assess the changes in meibomian glands in scleral lens wearers over one year of lens wear. Methods: Infrared meibography was obtained from the upper eyelid using the Cobra fundus camera in forty-three volunteers (34.2 ± 10.1 years). Meibographies were classified into 3 groups: Group 1 = good subjective gland visibility and gland drop-out < 1/3 of the total area; Group 2 = low visibility and gland drop-out < 1/3; and Group 3 = low visibility and gland drop-out > 1/3. Meibomian gland visibility metrics were then calculated using the developed algorithm from the pixel intensity values of meibographies. Repeatability of new metrics and their correlations with gland drop-out were assessed. Meibographies and ocular symptoms were also assessed after 1 year of scleral lens wear in 29 subjects. Results: Gland drop-out percentage was not statistically different between groups 1 and 2 (p = 0.464). Nevertheless, group 1 showed higher grey pixel intensity values than the other groups. Statistically significant correlations were found between gland visibility metrics and gland drop-out percentage. Repeatability was acceptable for all metrics, coefficient of variation achieving values between 0.52 and 3.18. While ocular symptoms decreased with scleral lens wear (p < 0.001), no statistically significant differences were found in gland drop-out percentage (p = 0.157) and gland visibility metrics (p > 0.217). Conclusions: The proposed method can assess meibomian gland visibility in an objective and repeatable way. Scleral lens wear appears to not adversely affect meibomian gland drop-out and visibility while might improve dry eye symptoms after one year of lens wear. These preliminary results should be confirmed with a control group.

Repeatability and reproducibility of Cobra HD fundus camera meibography in young adults with and without symptoms of dry eye

“…Alternatively, the Cobra HD (http://csoitalia.it) nonmydriatic digital fundus camera 11 uses Phoenix semi‐automated software, which is specially designed for MG analysis 14 and employs a ‘vectorised’ tool with Beziers curves that adapt rapidly to the eyelid shape 12 . Thus, the Phoenix semi‐automated software is both quicker and better than the rough segmentation obtained with ImageJ free‐hand selection 12 .…”

“…11 Despite this, the OCULUS Keratograph 5 M has been reported to have good interexaminer reproducibility (IER; mean difference between examiners of 0.08 ± 0.55 and 0.13 ± 0.50 grade units in two separate sessions, respectively) with low within-subject variability (WSV; 95% limits of agreement for two different examiners of −1.02 to +1.10 and −1.27 to +1.09 grade units, respectively). 13 Alternatively, the Cobra HD (csoit alia.it) nonmydriatic digital fundus camera 11 uses Phoenix semi-automated software, which is specially designed for MG analysis 14 and employs a 'vectorised' tool with Beziers curves that adapt rapidly to the eyelid shape. 12 Thus, the Phoenix semi-automated software is both quicker and better than the rough segmentation obtained with ImageJ free-hand selection.…”

Meibomian gland dysfunction and dry eye symptoms of fitted and over the counter contact lens wearers compared with non-contact lens wearing controls

“…[7] Despite this, the OCULUS Keratograph 5M has been reported to have good inter-examiner reproducibility (IER; mean difference between examiners of 0.08± 0.55 and 0.13 ± 0.50 grade units in two separate sessions, respectively) with low within-subject variability (WSV; 95% limits of agreement for two different examiners of −1.02 to +1.10 and −1.27 to +1.09 grade units, respectively). [2] Alternatively, the Cobra HD (csoitalia.it) nonmydriatic digital fundus camera [7] uses Phoenix semi-automated software, which is specially designed for MG analysis [8] and employs a 'vectorised' tool with Beziers curves that adapt rapidly to the eyelid shape. [9] Thus, the Phoenix semiautomated software is both quicker and better than the rough segmentation obtained with ImageJ free-hand selection.…”

“…The Cobra HD (csoitalia.it) non-mydriatic digital fundus camera [7] with Phoenix semi-automated software that is specially designed for MG analysis, [8] employs a 'vectorised' tool with Beziers curves that adapts rapidly to the eyelid shape. [9] The Phoenix software requires that the examiner mark the area of the eyelids and the location where the glands are observed, while the software subsequently calculates the percentage of MG loss.…”

“…The mean MG loss in the upper eyelid was higher in soft CL wearers compared with the control group (p=0.045). García-Marqués et al[8] assessed the effects of soft CL wear and the wear-duration on the tear film, MG loss of the upper eyelids, and MG appearance in the upper eyelids. They showed that CL wearers have significantly more MG loss in the upper eyelid compared with non-CL wearing controls (p<0.001).…”