Quantitative analysis of morphological and functional features in Meibography for Meibomian Gland Dysfunction: Diagnosis and Grading

Meibomian Gland Dysfunction (MGD) is a common ocular surface disease and a major cause of dry eye, associated with various ocular and systemic conditions, and negatively affecting patients' quality of life. Current diagnostic methods often rely on the experience of the clinician, with issues such as difficulty in quantification and observer variability, and some patients find it challenging to cooperate with squeezing tests. Non-contact meibomian gland photography allows for direct observation of the glands but lacks an automated quantification method.

Research Methods:

From January 1, 2019, to December 31, 2020, data were collected from 256 patients with dry eye symptoms and 56 healthy volunteers. Ocular examinations were conducted, and all procedures were approved by an ethics committee and followed the Declaration of Helsinki. All participants provided informed consent.

The 256 symptomatic patients were divided into the MGD group (195 patients) and the symptomatic non-MGD group (61 patients) based on symptoms and meibomian gland secretion criteria, with MGD patients further assessed for severity.

Multiple clinical assessments, including the Ocular Surface Disease Index (OSDI) questionnaire and corneal fluorescein staining, were performed, with one eye randomly selected for analysis.

A self-developed analyzer was used to analyze meibomian gland images, extracting parameters such as Gland Area (GA), Diameter Deformation Index (DI), Tortuosity Index (TI), and Signal Intensity (SI).

Variance analysis and chi-square tests were performed to analyze the relationship between the parameters and MGD, as well as to evaluate diagnostic and grading performance.

Research Results:

The MGD group showed significant differences in multiple indices such as OSDI and tear break-up time compared to the normal and symptomatic non-MGD groups. The MGD group had lower GA and TI, higher DI, and different SI frequency distributions compared to the other two groups.

Binary logistic regression indicated that an increase in DI and low SI were positively correlated with MGD, while increases in TI and GA were negatively correlated with MGD. The combination of DI, TI, GA, and SI had an AUC-ROC of 0.82 for identifying MGD.

Meibomian gland morphological parameters were correlated with the Meibomian Gland Dysfunction Severity Score (MGYSS) and the severity of MGD. DI had a higher AUC-ROC in early MGD identification, while GA was higher in the mid to late stages. The combination of DI, TI, and GA provided the highest AUC-ROC across all stages.

Conclusion:

Gland Area Ratio, Diameter Deformation, Tortuosity, and Signal Intensity can serve as biomarkers for MGD diagnosis and grading. The DI-TI-GA-SI combination (figure 1) offers good discrimination for identifying MGD patients, and the DI-TI-GA combination shows high accuracy in distinguishing MGD severity. Automated quantitative analysis provides a non-invasive, rapid, and easy-to-operate method for MGD diagnosis and grading.

figure 1: (A) the appearance of the lid margins and meibomian orifices, (B) the Meibography images with automatically segmented boundaries of the everted tarsal conjunc- tiva area as the region of interest (ROI), and (C) the segmented meibomian glands of representative subjects. The sequence from 1 to 4 is based on the current stage of MGs pathological progression from healthy to severe. Examples of the targeted individual features of tortuosity (yellow arrow), dilation & distortion (red arrow), and ghost glands with a declined signal index (blue arrow), an increased signal index (green arrow), and atrophy or drop out (white arrow) are marked with arrows. The acquiredmulti-parametric analysis results of the four representative Meibography images are shown in the table below the figure. DI: diameter deformation index; TI: tortuosity index SI: signal index; SD: standard deviation; GA: gland area ratio, which is the ratio of all meibomian gland area to the total analysis area.

journal homepage: https://www.journals.elsevier.com/eclinicalmedicine