Clustering-Based Analysis of Meibomian Gland Morphology for Automated Assessment of Meibomian Gland Dysfunction Using the MGD-1k Dataset
DOI:
https://doi.org/10.52783/jns.v14.1777Keywords:
Meibomian gland dysfunction, clustering analysis, MGD-1k dataset, k-means clusteringAbstract
Purpose: To develop a clustering-based framework for automated categorization of Meibomian gland morphology, addressing subjectivity in Meibomian Gland Dysfunction (MGD) diagnosis.
Methods: The MGD-1k dataset of 1,000 infrared meibography images was analyzed using k-means clustering to classify gland areas into three groups: low (Cluster 1), medium (Cluster 2), and high (Cluster 3). Pixel-based segmentation determined white pixel counts as a surrogate for gland area. Statistical validation, including ANOVA, was performed, and results were visualized with scatter plots, bar charts, and box plots.
Results: Clustering identified 417, 358, and 225 images in Clusters 1, 2, and 3, comprising 41.7%, 35.8%, and 22.5% of the dataset, respectively. Mean pixel counts were 39,123.99 (SD = 4,213.89), 23,960.30 (SD = 5,930.49), and 54,413.95 (SD = 6,378.15). ANOVA confirmed significant inter-cluster differences (p < 0.0001).
Conclusions: This framework objectively quantifies gland morphology, enabling severity stratification. The approach offers scalable diagnostic support and could be integrated with clinical tools or AI models. Future validation on diverse datasets is needed to confirm its broader applicability.
Downloads
Metrics
References
Tong L, Lim L, Tan D, Heng WJ, Lim J, Chan C, et al. Assessment and Management of Dry Eye Disease and Meibomian Gland Dysfunction: Providing a Singapore Framework. Asia-Pac J Ophthalmol [Internet]. 2021 Nov 1 [cited 2025 Jan 22];10(6):530–41. Available from: https://www.sciencedirect.com/science/article/pii/S216209892300052X
Chen H, McCann P, Lien T, Xiao M, Abraham AG, Gregory DG, et al. Prevalence of dry eye and Meibomian gland dysfunction in Central and South America: a systematic review and meta-analysis. BMC Ophthalmol [Internet]. 2024 Jan 31 [cited 2025 Jan 22];24(1):50. Available from: https://doi.org/10.1186/s12886-023-03249-w
Narang P, Donthineni PR, D’Souza S, Basu S. Evaporative dry eye disease due to meibomian gland dysfunction: Preferred practice pattern guidelines for diagnosis and treatment. Indian J Ophthalmol [Internet]. 2023 Apr [cited 2025 Jan 22];71(4):1348. Available from: https://journals.lww.com/ijo/fulltext/2023/04000/Evaporative_dry_eye_disease_due_to_meibomian_gland.38.aspx
Foulks GN. The Correlation Between the Tear Film Lipid Layer and Dry Eye Disease. Surv Ophthalmol [Internet]. 2007 Jul 1 [cited 2025 Jan 22];52(4):369–74. Available from: https://www.sciencedirect.com/science/article/pii/S0039625707000574
Jester JV, Parfitt GJ, Brown DJ. Meibomian gland dysfunction: hyperkeratinization or atrophy? BMC Ophthalmol [Internet]. 2015 Dec 17 [cited 2025 Jan 22];15(1):156. Available from: https://doi.org/10.1186/s12886-015-0132-x
Fineide F, Arita R, Utheim TP. The role of meibography in ocular surface diagnostics: A review. Ocul Surf [Internet]. 2021 Jan 1 [cited 2025 Jan 22];19:133–44. Available from: https://www.sciencedirect.com/science/article/pii/S1542012420300823
Lee JM, Jeon YJ, Kim KY, Hwang KY, Kwon YA, Koh K. Ocular surface analysis: A comparison between the LipiView® II and IDRA®. Eur J Ophthalmol [Internet]. 2021 Sep 1 [cited 2025 Jan 22];31(5):2300–6. Available from: https://doi.org/10.1177/1120672120969035
Lee Y, Hyon JY, Jeon HS. Characteristics of dry eye patients with thick tear film lipid layers evaluated by a LipiView II interferometer. Graefes Arch Clin Exp Ophthalmol [Internet]. 2021 May 1 [cited 2025 Jan 22];259(5):1235–41. Available from: https://doi.org/10.1007/s00417-020-05044-5
Lee SM, Chung SJ, Lew H. Evaluation of Tear Film Lipid Layer Thickness Measurements Obtained Using an Ocular Surface Interferometer in Nasolacrimal Duct Obstruction Patients. Korean J Ophthalmol [Internet]. 2018 Nov 23 [cited 2025 Jan 22];32(6):445–50. Available from: https://synapse.koreamed.org/articles/1109046
Zhang J, Wu Z, Sun L, Liu X hua. Function and Morphology of the Meibomian Glands Using a LipiView Interferometer in Rotating Shift Medical Staff. J Ophthalmol [Internet]. 2020 [cited 2025 Jan 22];2020(1):3275143. Available from: https://onlinelibrary.wiley.com/doi/abs/10.1155/2020/3275143
Zhao Y, Tan CLS, Tong L. Intra-observer and inter-observer repeatability of ocular surface interferometer in measuring lipid layer thickness. BMC Ophthalmol [Internet]. 2015 May 15 [cited 2025 Jan 22];15(1):53. Available from: https://doi.org/10.1186/s12886-015-0036-9
Arita R, Itoh K, Maeda S, Maeda K, Furuta A, Fukuoka S, et al. Proposed Diagnostic Criteria for Obstructive Meibomian Gland Dysfunction. Ophthalmology [Internet]. 2009 Nov 1 [cited 2025 Jan 22];116(11):2058-2063.e1. Available from: https://www.sciencedirect.com/science/article/pii/S0161642009004230
Diz-Arias E, Fernández-Jiménez E, Peral A, Gomez-Pedrero JA. A Comparative Study of Two Imaging Techniques of Meibomian Glands. Life [Internet]. 2023 Mar [cited 2025 Jan 22];13(3):791. Available from: https://www.mdpi.com/2075-1729/13/3/791
Wang J, Yeh TN, Chakraborty R, Yu SX, Lin MC. A Deep Learning Approach for Meibomian Gland Atrophy Evaluation in Meibography Images. Transl Vis Sci Technol [Internet]. 2019 Dec 18 [cited 2025 Jan 22];8(6):37. Available from: https://doi.org/10.1167/tvst.8.6.37
Zhang Z, Lin X, Yu X, Fu Y, Chen X, Yang W, et al. Meibomian Gland Density: An Effective Evaluation Index of Meibomian Gland Dysfunction Based on Deep Learning and Transfer Learning. J Clin Med [Internet]. 2022 Jan [cited 2025 Jan 22];11(9):2396. Available from: https://www.mdpi.com/2077-0383/11/9/2396
Saha RK, Chowdhury AMM, Na KS, Hwang GD, Eom Y, Kim J, et al. AI-based automated Meibomian gland segmentation, classification and reflection correction in infrared Meibography [Internet]. arXiv; 2022 [cited 2025 Jan 22]. Available from: http://arxiv.org/abs/2205.15543
Saha RK, Chowdhury AMM, Na KS, Hwang GD, Eom Y, Kim J, et al. Automated quantification of meibomian gland dropout in infrared meibography using deep learning. Ocul Surf [Internet]. 2022 Oct 1 [cited 2025 Jan 22];26:283–94. Available from: https://www.sciencedirect.com/science/article/pii/S1542012422000519
Arthur D, Vassilvitskii S. k-means++: the advantages of careful seeding. In: Proceedings of the eighteenth annual ACM-SIAM symposium on Discrete algorithms. USA: Society for Industrial and Applied Mathematics; 2007. p. 1027–35. (SODA ’07).
Lloyd S. Least squares quantization in PCM. IEEE Trans Inf Theory [Internet]. 1982 Mar [cited 2025 Jan 22];28(2):129–37. Available from: https://ieeexplore.ieee.org/document/1056489
Krishnan A, Salman Sarkar Md, Badavath L. Machine Learning Applications in Retinopathy of Prematurity Diagnosis Using the ROP Retinal Image Dataset. J Neonatal Surg [Internet]. 2025 Feb 7 [cited 2025 Feb 18];14(1S):820–7. Available from: https://www.jneonatalsurg.com/index.php/jns/article/view/1607
Sarafrazi S, Fayaz S, Reisdorf S, Shih KC, Yin Wu P. Harnessing the Power of Bayesian Neural Networks for Annotator Consensus Refinement to Enhance Meibomian Gland Dysfunction Classification. In: 2024 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) [Internet]. 2024 [cited 2025 Jan 22]. p. 5999–6007. Available from: https://ieeexplore.ieee.org/document/10821833
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
You are free to:
- Share — copy and redistribute the material in any medium or format
- Adapt — remix, transform, and build upon the material for any purpose, even commercially.
Terms:
- Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
- No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.
