Ehsan Zare Bidaki, Parvaneh Najafizadeh, Paul J Murphy
ThermOcular can quantify OST dynamics and can demonstrate expected post-blink cooling. It can analyse CR data for separate corneal ROI and over different post-blink time periods. These findings demonstrate the feasibility of ThermOcular for research use and support future studies to explore OST CR as an objective, non-invasive marker of tear film behaviour.
PURPOSE: To demonstrate proof-of concept of a dual camera (thermal/visible) imaging system (ThermOcular) for dynamic tracking of ocular surface temperature (OST) in healthy eyes.
METHODS: 25 healthy participants were recruited (16 females, 9 males; mean ± SD age, 35.6 ± 11.4 years) One randomly selected eye of each participant was imaged using ThermOcular. Each participant made two complete blinks and then avoided blinking (MBI) while fixating on a primary gaze target. Image processing algorithms of the two video streams by ThermOcular extracted the corneal OST data. Baseline temperature immediately after eye opening (t₀), and cooling rate (CR, °C/s) during the post-blink period (3, 4, 5, 6, 7, 8 s from eye opening (CRx)) and maximum inter-blink interval (CRMBI) were analysed for six corneal regions of interest (ROI) (entire, central, superior, inferior, nasal, temporal). Normality of data was assessed using Shapiro-Wilk and appropriate tests applied. Within ROIs, temporal effects on CR were analysed using Friedman tests with Wilcoxon signed-rank post-hoc comparisons. CR between ROI were assessed using rank-based two-way repeated-measures analysis.
RESULTS: Mean ± SD baseline OST (t₀) across all ROIs was 34.0 ± 0.92 °C, decreasing to 33.59 ± 1.11 °C at tMBI, with minimal spatial variation across ROIs (Range: 0.24 ± 0.09 °C). Mean CR were negative over all time periods, becoming progressively less negative over time (e.g., CR3: -0.065, CR6: -0.046, CR8: -0.038 °C/s). Mean ± SD MBI duration was 20.6 ± 8.4 s; mean CRMBI -0.026 °C/s. All ROI had significant changes in CR (p < 0.001), indicating time-dependent cooling. CR for central and temporal ROI were cooler than nasal ROI, but were not statistically significant.
CONCLUSIONS: ThermOcular can quantify OST dynamics and can demonstrate expected post-blink cooling. It can analyse CR data for separate corneal ROI and over different post-blink time periods. These findings demonstrate the feasibility of ThermOcular for research use and support future studies to explore OST CR as an objective, non-invasive marker of tear film behaviour.