Repeatability of grading meibomian gland dropout using two infrared systems

Area of Science:

• Ophthalmology
• Medical Imaging
• Biomedical Engineering

Background:

• Meibomian gland (MG) dysfunction is a common cause of dry eye disease.
• Meibography is a non-invasive technique to visualize and assess MG dropout.
• Standardized grading scales are crucial for reliable assessment of MG dropout.

Purpose of the Study:

• To evaluate the interobserver and intraobserver repeatability of meibomian gland dropout grading using OCULUS Keratograph 4 (K4) and 5M (K5M) devices.
• To compare the results obtained from the K4 and K5M devices for semiobjective quantification of MG dropout.
• To explore the correlation between meibography scores and clinical signs of dry eye.

Main Methods:

• 40 participants underwent meibography using K4 and K5M devices, with images taken thrice per eye.
• Two observers graded meibomian gland dropout using a four-point scale on two separate occasions.
• Semiobjective percentage dropout was quantified using ImageJ, and a seven-point scale was also evaluated.

Main Results:

• Both K4 and K5M showed good interobserver and intraobserver repeatability for MG dropout grading, especially with a seven-point scale (CCC up to 0.95).
• However, there was significant disagreement (21.8% coefficient of repeatability) in percentage dropout measurements between K4 and K5M images.
• Meibography scores did not correlate with clinical signs but showed a high correlation with percentage dropout.

Conclusions:

• The OCULUS K4 and K5M devices demonstrate similar inter- and intraobserver repeatability for grading meibomian gland dropout.
• The significant disagreement in percentage dropout suggests K4 and K5M are not interchangeable for this specific measurement.
• A finer grading scale may improve the detection of changes in meibomian gland dropout.