Ocular surface parameters and the recording of tear film particles’ spreading post-blink were assessed in eighty-one healthy volunteers (43.7 ± 27.0 years) using Keratograph 5M. The developed software automatically decomposed the video into frames to manually track particles’ position for 1.75 seconds after a blink. The following tear film-dynamic metrics were automatically calculated: mean, median, maximum and minimum particles’ speed at different times after blinking and time for particle speed to decrease to < 1.20 mm/second. Repeatability of each tear film-dynamic metric and its correlations with ocular surface signs and symptoms were analyzed. Binomial logistic regression was performed to assess the predictability of new metrics to ocular parameters.
Repeatability tended to be lower just after blinking (variability of 12.24 %), whereas the metrics from 0.5 seconds onwards had acceptable repeatability (variability below 10 %). Tear film-dynamic metrics correlated positively with Non-Invasive Break-Up Time (NIKBUT) while negatively with meibomian gland drop-out. Binomial logistic regression analysis revealed that tear film-dynamic metrics were able to predict NIKBUT. Nevertheless, no statistically significant association was found with gland drop-out. This means that higher particle speed is related to larger NIKBUT. The metric “time for particle speed to decrease to < 1.20 mm/second" can be considered the best metric to assess the quality of the tear film, since it was more strongly correlated with NIKBUT (r=0.42, p=0.004), it was more strongly associated in the binomial logistic regression analysis with NIKBUT and showed good repeatability (variability = 5.49 %).
Tear film-dynamic metrics are emerging homeostasis parameters for assessing indirectly the tear film quality in natural conditions with acceptable repeatability.