Latest developments in meibography: A review

Investigating Risk Factors for Meibomian Gland Dysfunction and Loss Among Young Medical Trainees

PURPOSE

To determine risk factors for meibomian gland disease and associated structural abnormalities in meibography among young medical trainees.

METHODS

This study included 84 medical students and residents younger than 45 years. All participants completed an ocular history and lifestyle questionnaire and the standardized patient evaluation of eye dryness (SPEED) II questionnaire. Meibomian gland (MG) dropout and structural changes were evaluated using meibography and scored by 2 graders using meiboscores. Statistical analysis aimed to identify MG loss risk factors.

RESULTS

Eighty-four individuals participated, and 168 meibography images were reviewed. Previous contact lens (CL) use (n = 88) demonstrated increased MG loss (P = 0.007). Correlation analysis revealed positive associations between MG loss and both frequency (Spearman r = 0.212, P = 0.003) and duration (Spearman r = 0.271, P <0.001) of CL use. Eye makeup users (n = 112) showed greater MG loss than nonusers (n = 56, P = 0.030), particularly eyeliner (n = 102) and eyeshadow (n = 100) users showing increased meiboscores (P = 0.020 and P = 0.040, respectively). Self-reported screen time and SPEED II scores did not correlate with meiboscores (P >0.05).

CONCLUSIONS

In a young trainee population, where age-related factors are reduced, previous CL use and eye makeup use are significant risk factors for MG loss. Frequency and length of CL wear affect MG dropout severity. Eye makeup usage also affected MG loss. Highlighting the incongruence of symptoms to signs, SPEED II scores showed no relationship with MG structural integrity.

Treatment of Meibomian Gland Dysfunction by Classical Eyelid Hygiene Measures With and Without Additional Lipid Substitution for Tear Film Stabilization

OBJECTIVE

This study aimed to document the treatment success of eyelid hygiene with liposomal suspension using new diagnostic tools and to determine whether additional lipid substitution provided measurable benefits in meibomian gland dysfunction.

METHODS

A single-center controlled, partially single masked study was conducted. Group A used eyelid hygiene only, whereas group B additionally applied a liposomal eye spray. Subjective perception using the Ocular Surface Disease Index (OSDI) questionnaire, measurement of tear film osmolarity, noninvasive tear film break-up time, assessment of the lipid layer, meibography, inspection of lid margins, assessment of the meibomian glands, and measurement of matrix-metallopeptidase-9 were collected at baseline and after 12 weeks.

RESULTS

Eighty-two patients were included and randomized into two groups. Both groups showed a decrease in OSDI score (P<0.001), an increase in lipid layer thickness (P<0.001), significant improvement in the degree of loss of meibomian glands (P<0.001), in the damage pattern of the eyelid margins (P<0.001), in the quality of meibomian gland secretion, and in matrix-metallopeptidase-9 after 12 weeks.

CONCLUSIONS

Our study confirms the success of treatment with eyelid hygiene measures using a liposomal suspension. The additional benefit of lipid substitution was not significant after 12 weeks of treatment.

Advances in artificial intelligence for meibomian gland evaluation: A comprehensive review

Meibomian gland dysfunction (MGD) is increasingly recognized as a critical contributor to evaporative dry eye, significantly impacting visual quality. With a global prevalence estimated at 35.8 %, it presents substantial challenges for clinicians. Conventional manual evaluation techniques for MGD face limitations characterized by inefficiencies, high subjectivity, limited big data processing capabilities, and a dearth of quantitative analytical tools. With rapidly advancing artificial intelligence (AI) techniques revolutionizing ophthalmology, studies are now leveraging sophisticated AI methodologies--including computer vision, unsupervised learning, and supervised learning--to facilitate comprehensive analyses of meibomian gland (MG) evaluations. These evaluations employ various techniques, including slit lamp examination, infrared imaging, confocal microscopy, and optical coherence tomography. This paradigm shift promises enhanced accuracy and consistency in disease evaluation and severity classification. While AI has achieved preliminary strides in meibomian gland evaluation, ongoing advancements in system development and clinical validation are imperative. We review the evolution of MG evaluation, juxtapose AI-driven methods with traditional approaches, elucidate the specific roles of diverse AI technologies, and explore their practical applications using various evaluation techniques. Moreover, we delve into critical considerations for the clinical deployment of AI technologies and envisages future prospects, providing novel insights into MG evaluation and fostering technological and clinical progress in this arena.

Effect of Heating and Massaging of Meibomian Glands on Their Imaging

Background and Objectives: Infrared light is used to image the Meibomian glands through their thermal profile. This study aimed to investigate the effects of a combination of heating and an eyelid massage on Meibomian gland visibility and tear film parameters. Materials and Methods: Twenty-four participants (26 ± 6.9 years) were enrolled in this prospective study, which involved imaging the Meibomian glands of both the lower and upper eyelid and assessing the non-invasive breakup time (NIBUT), tear meniscus height (TMH), and blink rate (using the CA-800, Topcon) at baseline after five minutes of eyelid warming followed by a five-minute eyelid massage. The second session, which was randomised in sequence, repeated the same measurements but without the inclusion of any eyelid warming or massage as the control condition. Results: While there was no change in lower lid Meibomian gland appearance as a result of eyelid heating, eyelid massage, or multiple lid eversion (median 2.0, range 0.0 to 4.0; p = 0.782), there was a change in upper lid appearance 5 min after heating and lid massage (p = 0.025), but again, multiple lid eversion had no effect (p > 0.05). The NIBUT decreased on second lid eversion (p = 0.049), although this was not evident on the third lid eversion (p = 0.090). The effect on NIBUT was also apparent with heating (p = 0.034 immediately after) but was sustained with 5 min of eyelid massage (p = 0.031). The TMH increased with heating (p < 0.001), and this effect was sustained with 5 min of eyelid massage (p = 0.011), but there was no lid eversion effect (p > 0.05). The blink rate was unaffected by heating, eyelid massage, or multiple eversions of the eyelids (median 24 blinks/min, range 8 to 59 blinks/min; p = 0.61). Conclusions: Eyelid warming can increase the visibility of the Meibomian glands, although this effect was only observed with upper lid imaging and the effect dissipated after 5 min of eyelid massage. Warming and massage also disrupt the tear film, as does multiple lid eversion, emphasising the need to use the least invasive tear film assessment techniques first.

Dual-Mode Imaging System for Early Detection and Monitoring of Ocular Surface Diseases

The global prevalence of ocular surface diseases (OSDs), such as dry eyes, conjunctivitis, and subconjunctival hemorrhage (SCH), is steadily increasing due to factors such as aging populations, environmental influences, and lifestyle changes. These diseases affect millions of individuals worldwide, emphasizing the importance of early diagnosis and continuous monitoring for effective treatment. Therefore, we present a deep learning-enhanced imaging system for the automated, objective, and reliable assessment of these three representative OSDs. Our comprehensive pipeline incorporates processing techniques derived from dual-mode infrared (IR) and visible (RGB) images. It employs a multi-stage deep learning model to enable accurate and consistent measurement of OSDs. This proposed method has achieved a 98.7% accuracy with an F1 score of 0.980 in class classification and a 96.2% accuracy with an F1 score of 0.956 in SCH region identification. Furthermore, our system aims to facilitate early diagnosis of meibomian gland dysfunction (MGD), a primary factor causing dry eyes, by quantitatively analyzing the meibomian gland (MG) area ratio and detecting gland morphological irregularities with an accuracy of 88.1% and an F1 score of 0.781. To enhance convenience and timely OSD management, we are integrating a portable IR camera for obtaining meibography during home inspections. Our system demonstrates notable improvements in expanding dual-mode image-based diagnosis for broader applicability, effectively enhancing patient care efficiency. With its automation, accuracy, and compact design, this system is well-suited for early detection and ongoing assessment of OSDs, contributing to improved eye healthcare in an accessible and comprehensible manner.

Global research of artificial intelligence in eyelid diseases: A bibliometric analysis

Purpose

To generate an overview of global research on artificial intelligence (AI) in eyelid diseases using a bibliometric approach.

Methods

All publications related to AI in eyelid diseases from 1900 to 2023 were retrieved from the Web of Science (WoS) Core Collection database. After manual screening, 98 publications published between 2000 and 2023 were finally included. We analyzed the annual trend of publication and citation count, productivity and co-authorship of countries/territories and institutions, research domain, source journal, co-occurrence and evolution of the keywords and co-citation and clustering of the references, using the analytic tool of the WoS, VOSviewer, Wordcloud Python package and CiteSpace.

Results

By analyzing a total of 98 relevant publications, we detected that this field had continuously developed over the past two decades and had entered a phase of rapid development in the last three years. Among these countries/territories and institutions contributing to this field, China was the most productive country and had the most institutions with high productivity, while USA was the most active in collaborating with others. The most popular research domains was Ophthalmology and the most productive journals were Ocular Surface. The co-occurrence network of keywords could be classified into 3 clusters respectively concerned about blepharoptosis, meibomian gland dysfunction and blepharospasm. The evolution of research hotspots is from clinical features to clinical scenarios and from image processing to deep learning. In the clustering analysis of co-cited reference network, cluster "0# deep learning" was the largest and latest, and cluster "#5 meibomian glands visibility assessment" existed for the longest time.

Conclusions

Although the research of AI in eyelid diseases has rapidly developed in the last three years, there are still gaps in this area. Our findings provide researchers with a better understanding of the development of the field and a reference for future research directions.

Temporal variations in meibomian gland structure-A pilot study

PURPOSE

To investigate whether there is a measurable change in meibomian gland morphological characteristics over the course of a day (12 h) and over a month.

METHODS

The study enrolled 15 participants who attended a total of 11 study visits spanning a 5-week period. To assess diurnal changes in meibomian glands, seven visits were conducted on a single day, each 2 h apart. For monthly assessment, participants attended an additional visit at the same time of the day every week for three consecutive weeks. Meibography using the LipiView® II system was performed at each visit, and meibomian gland morphological parameters were calculated using custom semi-automated software. Specifically, six central glands were analysed for gland length ratio, gland width, gland area, gland intensity and gland tortuosity.

RESULTS

The average meibomian gland morphological metrics did not exhibit significant changes during the course of a day or over a month. Nonetheless, certain individual gland metrics demonstrated notable variation over time, both diurnally and monthly. Specifically, meibomian gland length ratio, area, width and tortuosity exhibited significant changes both diurnally and monthly when assessed on a gland-by-gland basis.

CONCLUSIONS

Meibomian glands demonstrated measurable structural change over short periods of time (hours and days). These results have implications for innovation in gland imaging and for developing precision monitoring of gland structure to assess meibomian gland health more accurately.

Evaluation of Meibomian gland structure and appearance after therapeutic Meibomian gland expression

CLINICAL RELEVANCE

Evaluating how Meibomian glands can change in appearance has the potential to advance the understanding of Meibomian gland health and may lead to enhanced diagnosis and therapy.

BACKGROUND

This work aimed to investigate Meibomian gland appearance after therapeutic Meibomian gland expression.

METHODS

Fifteen subjects attended three study visits over a two-week period. Meibography was performed before and after therapeutic Meibomian gland expression, the following day, and 2 weeks after expression. Six central glands were used to calculate Meibomian gland morphological parameters such as gland length ratio, gland width, gland area, gland tortuosity, and gland contrast. A custom semi-automated image analysis software was used to calculate Meibomian gland metrics. Furthermore, a high-resolution imaging system was developed to capture clear images of the Meibomian glands, free of any artefacts, which were used for precise calculations of Meibomian gland contrast.

RESULTS

The expression procedure had a significant impact on Meibomian gland contrast and length ratio immediately afterwards. The least square mean difference (95% CI) from baseline for Michelson contrast was -0.006 (-0.010, -0.001) and -1.048 (-2.063, -0.033) for simple contrast. The least square mean ratio of the gland length ratio immediately after the expression to baseline was 0.758 (0.618, 0.931).

CONCLUSIONS

Following therapeutic expression, Meibomian glands exhibit reduced brightness and length. However, within 24 h, they appear to recover and return to their baseline state, indicating a relatively short recovery time. This sheds light on whether meibography is solely focused on capturing gland structure or if it also captures acinar activity. The hyperreflective properties of lipids suggest that the decrease in contrast observed after expression could be attributed to a reduction in the visualisation of acini activity. A decrease in Meibomian gland length ratio implies that the loss of gland structure following treatment may be indicative of a temporary structural alteration.

2022 Glenn A. Fry Award lecture: Enhancing clinical assessment for improved ophthalmic management

ABSTRACT

Detailed clinical assessment is critical to allow sensitive evaluation of the eye and its management. As technology advances, these assessment techniques can be adapted and refined to improve the detection of pathological changes of ocular tissue and their impact on visual function. Enhancements in optical medical devices including spectacle, contact, and intraocular lenses have allowed for a better understanding of the mechanism and amelioration of presbyopia and myopia control. Advancements in imaging technology have enabled improved quantification of the tear film and ocular surface, informing diagnosis and treatment strategies. Miniaturized electronics, large processing power, and in-built sensors in smartphones and tablets capacitate more portable assessment tools for clinicians, facilitate self-monitoring and treatment compliance, and aid communication with patients. This article gives an overview of how technology has been used in many areas of eye care to improve assessments and treatment and provides a snapshot of some of my studies validating and using technology to inform better evidence-based patient management.

Contact lens wear and follow-up and its association with signs and symptoms of meibomian gland dysfunction

INTRODUCTION

This study investigated the equivocal association between contact lens (CL) wear and meibomian gland dysfunction (MGD) by comparing the morphological, functional and subjective outcomes of CL wearers versus control, non-CL wearers. CL wearers were examined as two cohorts based on the annual attendance of follow-up visits (FLU-attended these visits, whereas non-FLU did not).

METHODS

Habitual logMAR visual acuity, invasive and non-invasive tear break-up time, Schirmer test, Efron grading scales, meibum quality score (MQS), meibum expressibility score (MES), meibomian gland (MG) loss, lid margin abnormalities and subjective dry eye (DE) symptoms were assessed.

RESULTS

Of the 128 participants, 31 were in the FLU cohort, 43 were in the non-FLU cohort and 54 were controls (mean ages: 22.2 ± 3.1, 23.0 ± 4.6 and 22.3 ± 3.5, respectively). Non-FLU CL wearers had more symptoms than controls (3.7 ± 2.4 vs. 2.3 ± 2.1, p < 0.01). Morphologically, FLU (16.9 ± 8.8%, p = 0.02) and non-FLU (18.6 ± 11.3%, p = 0.001) had more MG loss than controls (11.2 ± 6.8%). Functionally, FLU (0.6 ± 0.7, p = 0.01) and non-FLU (0.8 ± 0.9, p = 0.001) had worse MES than controls (0.2 ± 0.5). FLU and non-FLU were both associated with corneal staining (odds ratio [OR] = 3.42, 95% CI: 1.16-10.11, p = 0.03 and OR = 5.23, 95% CI: 1.89-14.48, p = 0.001, respectively) and MG loss (OR = 10.47, 95% CI: 1.14-96.29, p = 0.04 and OR = 16.63, 95% CI: 1.96-140.86, p = 0.01, respectively). Non-FLU CL wear was also associated with abnormal MQS (OR = 12.87, 95% CI: 1.12-148.41, p = 0.04), conjunctival staining (OR = 12.18, 95% CI: 3.66-40.51, p < 0.001) and lid margin telangiectasia (OR = 3.78, 95% CI: 1.55-9.21, p = 0.003). MGD was three times more prevalent in CL wearers (12%) than in controls (4%).

CONCLUSIONS

Both CL-wearing cohorts demonstrated significantly more MG abnormalities than controls though the difference was not clinically significant. Non-FLU CL wearers had more DE symptoms. Non-FLU CL wear is an independent predictor for more abnormalities than FLU CL wear, emphasising the importance of follow-ups.

A Deep Learning Approach for Meibomian Gland Appearance Evaluation

Purpose

To develop and evaluate a deep learning algorithm for Meibomian gland characteristics calculation.

Design

Evaluation of diagnostic technology.

Subjects

A total of 1616 meibography images of both the upper (697) and lower (919) eyelids from a total of 282 individuals.

Methods

Images were collected using the LipiView II device. All the provided data were split into 3 sets: the training, validation, and test sets. Data partitions used proportions of 70/10/20% and included data from 2 optometry settings. Each set was separately partitioned with these proportions, resulting in a balanced distribution of data from both settings. The images were divided based on patient identifiers, such that all images collected for one participant could end up only in one set. The labeled images were used to train a deep learning model, which was subsequently used for Meibomian gland segmentation. The model was then applied to calculate individual Meibomian gland metrics. Interreader agreement and agreement between manual and automated methods for Meibomian gland segmentation were also carried out to assess the accuracy of the automated approach.

Main Outcome Measures

Meibomian gland metrics, including length ratio, area, tortuosity, intensity, and width, were measured. Additionally, the performance of the automated algorithms was evaluated using the aggregated Jaccard index.

Results

The proposed semantic segmentation-based approach achieved average aggregated Jaccard index of mean 0.4718 (95% confidence interval [CI], 0.4680-0.4771) for the 'gland' class and a mean of 0.8470 (95% CI, 0.8432-0.8508) for the 'eyelid' class. The result for object detection-based approach was a mean of 0.4476 (95% CI, 0.4426-0.4533). Both artificial intelligence-based algorithms underestimated area, length ratio, tortuosity, widthmean, widthmedian, width10th, and width90th. Meibomian gland intensity was overestimated by both algorithms compared with the manual approach. The object detection-based algorithm seems to be as reliable as the manual approach only for Meibomian gland width10th calculation.

Conclusions

The proposed approach can successfully segment Meibomian glands; however, to overcome problems with gland overlap and lack of image sharpness, the proposed method requires further development. The study presents another approach to utilizing automated, artificial intelligence-based methods in Meibomian gland health assessment that may assist clinicians in the diagnosis, treatment, and management of Meibomian gland dysfunction.

Financial Disclosures

The authors have no proprietary or commercial interest in any materials discussed in this article.

Risk Factors for Meibomian Gland Disease Assessed by Meibography

Purpose

To elucidate risk factors for meibomian gland disease (MGD) and understand associated changes in meibography and in relation to ocular surface disease.

Patients and Methods

As part of the standard workup for ocular surface disease at a tertiary academic center, 203 patients received an ocular history and lifestyle questionnaire. The questionnaire included detailed inquiries about ocular health and lifestyle, including makeup use, cosmetic eyelid procedures, screen time, and contact lens habits. Subjects also took the standardized patient evaluation of eye dryness (SPEED) II questionnaire. Meibomian gland (MG) dropout and structural changes were evaluated on meibography and scored by three independent graders using meiboscores. Statistical analysis was conducted to identify significant risk factors associated with MG loss.

Results

This retrospective, cross-sectional study included 189 patients (378 eyes) with high-quality images for grading, and the average age was 67 years (77% female). Patients older than 45 years had significantly more dropout than younger patients (p < 0.01). Self-reported eye makeup use did not significantly impact MG loss. Patients with a history of blepharoplasty trended toward higher meiboscores, but the difference was not statistically significant. Self-reported screen time did not affect meiboscores. Contact lens use over 20 years was associated with significant MG loss (p < 0.05). SPEED II scores had no relationship to meiboscores (p = 0.75).

Conclusion

Older age is a significant risk factor for MG loss. Any contact lens use over 20 years also impacted MG dropout. Highlighting the incongruence of symptoms to signs, SPEED II scores showed no relationship to the structural integrity of MGs.

Unsupervised Learning Based on Meibography Enables Subtyping of Dry Eye Disease and Reveals Ocular Surface Features

Purpose

This study aimed to establish an image-based classification that can reveal the clinical characteristics of patients with dry eye using unsupervised learning methods.

Methods

In this study, we analyzed 82,236 meibography images from 20,559 subjects. Using the SimCLR neural network, the images were categorized. Data for each patient were averaged and subjected to mini-batch k-means clustering, and validated through consensus clustering. Statistical metrics determined optimal category numbers. Using a UNet model, images were segmented to identify meibomian gland (MG) areas. Clinical features were assessed, including tear breakup time (BUT), tear meniscus height (TMH), and gland atrophy. A thorough ocular surface evaluation was conducted on 280 cooperative patients.

Results

SimCLR neural network achieved clustering patients with dry eye into six image-based subtypes. Patients in different subtypes harbored significantly different noninvasive BUT, significantly correlated with TMH. Subtypes 1 and 5 had the most severe MG atrophy. Subtype 2 had the highest corneal fluorescent staining (CFS). Subtype 4 had the lowest TMH, whereas subtype 5 had the highest. Subtypes 3 and 6 had the largest MG areas, and the upper MG areas of a person's bilateral eyes were highly correlated. Image-based subtypes are related to meibum quality, CFS, and morphological characteristics of MG.

Conclusions

In this study, we developed an unsupervised neural network model to cluster patients with dry eye into image-based subtypes using meibography images. We annotated these subtypes with functional and morphological clinical characteristics.

Morphologic variants of Meibomian glands: age-wise distribution and differences between upper and lower eyelids

Purpose

To evaluate the distribution of various Meibomian gland morphologies across different age groups in healthy individuals.

Methods

The infrared meibographic morphologies of the Meibomian glands from the upper and lower eyelids of 236 healthy individuals (472 eyes; mean age 38.4 ± 17.5 years; 80 female participants: 156 male participants) were evaluated for their prevalence and differences across six decades of life, from 10 to 80 years. A linear mixed-effects modeling test was performed for statistical analysis.

Results

Of 14,452 glands, 8,830 (61%) glands were located in the upper eyelid. No significant differences in frequency were noted between different age groups for distorted, tortuous, hooked, overlapping, abnormal gap, fluffy areas, dropout (except for 51-60 vs. 10-20 years, P = 0.023), and thick and thin morphologies. Short glands were significantly more common in individuals aged over 30 years (P = 0.015), whereas moderately short and severely short glands were more common in the upper eyelids of individuals older than 50 years compared to those aged 10-20 years (P = 0.035). The frequency of distorted, hooked, tortuous, overlapping, and tadpole-shaped Meibomian glands was significantly higher in the upper eyelids than in the lower eyelids for all age groups. Dropout glands were more common in the lower eyelids of individuals younger than 50 years, but no difference was observed in the upper and lower eyelids of individuals over 50 years. Dropout (P = 0.006) and severely short glands (0.026) of the lower eyelid were associated with low non-invasive tear break-up time (NIBUT) values.

Conclusion

Various morphologic characteristics of the Meibomian glands that are considered abnormal can be present in healthy individuals, and only moderate to severely short glands display an increase in abnormal morphologic characteristics of the Meibomian glands with age.