Open Access Peer-reviewed Research Article

A Data-Driven Evaluation of ECD Measurement Techniques Across Traditional and AI-Based Modalities

Main Article Content

Manghe Fidelis Obi
Andy Officer
Shannon Schweitzer
Tarun Goswami corresponding author

Abstract

Accurate measurement of corneal endothelial cell density (ECD) is crucial in evaluating the viability of donor corneas for transplantation. The consistency of ECD measurements is critical for predicting post-transplant results and monitoring corneal health. However, measurement methods have evolved, moving from manual counting to more complex semi-automatic and fully automated systems, including AI-powered solutions. This study compares the accuracy, dependability, and efficiency of manual, semi-automated, and fully automated ECD measurement techniques. It investigates the degree of heterogeneity among techniques and evaluates their potential to improve clinical outcomes in corneal transplantation. The sample includes corneal data from 300 participants, 150 male and 150 female donors, who were divided into three groups based on the measurement method: manual, semi-automated, or fully automated. The study also examined the gender distribution to see whether there was any difference in results between male and female donor corneas. Manual counting has previously been notable for its variability due to operator expertise and calibration discrepancies, with mean ECD values ranging from 2146 to 2775 cells/mm² (p < 0.05). Semi-automated procedures, which combine manual input with software aid, enhance consistency. In the Cornea Preservation Time Study, eye banks reported a mean ECD of 2773 ± 300 cells/mm², while CIARC reported 2758 ± 388 cells/mm², with agreement limits ranging from [-644, 675] cells/mm² (p < 0.05). The AxoNet deep learning model had a mean absolute error (MAE) of 12.1 cells/mm² and an R² value of 0.948, making it the most accurate fully automated system. A separate study on AI-based detection of aberrant endothelium cells achieved an accuracy of 0.95, precision of 0.92, recall of 0.94, and F1 score of 0.93, and an AUC-ROC of 0.98 (p < 0.01). Fully automated AI-based methods surpass manual and semi-automated approaches in accuracy and consistency, significantly reducing time and labor. The findings highlight the importance of adopting AI-driven technologies to enhance diagnostic precision and efficiency in clinical settings. However, the need for standardized calibration procedures and high- quality image acquisition remains critical for reliable ECD measurement.

Keywords
corneal endothelial cell density (ECD), manual cell counting, semi-automated cell counting, fully automated cell counting, deep learning, image analysis, AI in ophthalmology

Article Details

Supporting Agencies
The successful completion of this research was made possible through the collective efforts of individuals and institutions whose contributions are deeply appreciated. Special thanks are extended to the faculty and research staff whose guidance and expertise have provided invaluable insights into the methodologies explored in this study. Their critical feedback and support have played a significant role in refining the technical and analytical aspects of this work. Gratitude is also expressed to the research community whose previous studies and innovations in endothelial cell density measurement have served as the foundation for this work. The continuous advancements in machine learning, deep learning, and AI-driven automation have significantly shaped the direction and execution of this study. A special acknowledgment is extended to Dr. Darryl Ahner, Dean of the College of Engineering and Computer Science, whose financial support enabled the successful completion of this project. The resources provided have facilitated access to computational tools, datasets, and analytical frameworks essential for conducting this research. Finally, appreciation is given to colleagues, peers, and family members for their encouragement and support throughout the research process. Their motivation and feedback have contributed to the successful realization of this work.
How to Cite
Obi, M. F., Officer, A., Schweitzer, S., & Goswami, T. (2025). A Data-Driven Evaluation of ECD Measurement Techniques Across Traditional and AI-Based Modalities. Research on Intelligent Manufacturing and Assembly, 4(2), 239-265. https://doi.org/10.25082/RIMA.2025.02.002

References

  1. Janson BJ, Alward WL, Kwon YH, et al. Glaucoma-associated corneal endothelial cell damage: A review. Survey of Ophthalmology. 2018, 63(4): 500-506. https://doi.org/10.1016/j.survophthal.2017.11.002
  2. Mannis MJ, Holland EJ. Cornea, E-Book. Elsevier Health Sciences, 2021.
  3. McCarey BE, Edelhauser HF, Lynn MJ. Review of Corneal Endothelial Specular Microscopy for FDA Clinical Trials of Refractive Procedures, Surgical Devices, and New Intraocular Drugs and Solutions. Cornea. 2008, 27(1): 1-16. https://doi.org/10.1097/ico.0b013e31815892da
  4. Bourne WM. Cell Density in Corneal Endothelium: Effect of Age and Measurement Method. Invest Ophthalmol Vis Sci. 1982, 22: 104-108.
  5. Laing RA, Sandstrom MM, Berrospi AR, et al. Changes in the corneal endothelium as a function of age. Experimental Eye Research. 1979, 29(4): 521-527.
  6. Mishima S. Clinical Investigations on the Corneal Endothelium. American Journal of Ophthalmology. 1982, 93(1): 1-29. https://doi.org/10.1016/0002-9394(82)90693-6
  7. Brown NAP, Bron AJ. Endothelial cell size, shape, and apical structure in the aging human cornea. British Journal of Ophthalmology. 1977, 6(11): 805-814.
  8. McCarey BE, Edelhauser HF. The effects of age on the corneal endothelium. Ophthalmology. 1981, 88(6): 649-660.
  9. Van Rossum G, Drake FL. Python 3 Reference Manual. Scotts Valley, CA: CreateSpace, 2009.
  10. McKinney W. Data Structures for Statistical Computing in Python. Proceedings of the 9th Python in Science Conference. Published online 2010: 56-61. https://doi.org/10.25080/majora-92bf1922-00a
  11. McLaren JW, Bourne WM. A computer-assisted technique for studying regular photomicrographs of the corneal endothelium. Cornea. 1991, 10(5): 445-453.
  12. Yee RW, Matsuda M, Kern JA. Computer-assisted endothelial morphometry in corneal endothelial images obtained by specular microscopy. American Journal of Ophthalmology‌. 1985, 99(6): 739-747.
  13. Smiddy WE, Bourne WM. The effects of long-term contact lens wear on the corneal endothelium. Archives of Ophthalmology‌. 1989, 107(10): 1421-1425.
  14. Hatou S, Miyashita H, Shimmura S, et al. Evaluation of the corneal endothelium using a newly developed noncontact specular microscope. Cornea. 2009, 28(4): 456-460.
  15. Murphy C, Alvarado J, Juster R, et al. Computerized morphometry of the corneal endothelium in long-term soft contact lens wearers. American Journal of Ophthalmology‌. 1984, 98(4): 478-486.
  16. Rivolta MW, Mantero M, Cereda M, et al. Automated analysis of corneal endothelial cell morphology using a novel software. PLoS ONE. 2017, 12(8): e0182585.
  17. Chui J, Girolamo ND, Wakefield D, et al. The Pathogenesis of Pterygium: Current Concepts and Their Therapeutic Implications. The Ocular Surface. 2008, 6(1): 24-43. https://doi.org/10.1016/s1542-0124(12)70103-9
  18. Yazdanpanah G, Jabbehdari S, Dana R. Corneal Endothelial Cell Dysfunction: Etiologies and Management. Curr Ophthalmol Rep. 2015, 3(4): 159-165.
  19. Benitez BA, Yee RW, Matsuda M, et al. Automated analysis of corneal endothelial cell images: comparison of methodologies. Cornea. 2010, 29(3): 287-294.
  20. Bourne WM. Cell Density in Corneal Endothelium: Effect of Age and Measurement Method. Invest Ophthalmol Vis Sci. 1982, 22: 104-108.
  21. Laing RA, Sandstrom MM, Berrospi AR, et al. Changes in the corneal endothelium as a function of age. Experimental Eye Research‌. 1979, 29(4): 521-527.
  22. Brown NAP, Bron AJ. Endothelial cell size, shape, and apical structure in the aging human cornea. British Journal of Ophthalmology. 1977, 61(11): 805-814.
  23. McCarey BE, Edelhauser HF. The effects of age on the corneal endothelium. Ophthalmology. 1981, 88(6): 649-660.
  24. Doughty MJ, Müller A, Zaman ML. Assessment of the Reliability of Human Corneal Endothelial Cell-Density Estimates Using a Noncontact Specular Microscope. Cornea. 2000, 19(2): 148-158. https://doi.org/10.1097/00003226-200003000-00006
  25. Kolluru C, Benetz BA, Joseph N, et al. Machine learning for segmenting cells in corneal endothelium images. Proceedings of Spie--the International Society for Optical Engineering. 2019, 10950: 109504G.
  26. Koprowski R, Wilczynski S, Wozniak M, et al. Computer-aided analysis of corneal endothelium images based on deep learning methods. Computerized Medical Imaging and Graphics‌. 2014, 44: 29–38.
  27. Modak C, Huang J, Maram J, et al. Automated endothelial cell density analysis with deep learning. Journal of Optometry‌. 2015, 12: 142–149.
  28. Price MO, Fairchild KM, Price FW. Comparison of Manual and Automated Endothelial Cell Density Analysis in Normal Eyes and DSEK Eyes. Cornea. 2013, 32(5): 567-573. https://doi.org/10.1097/ico.0b013e31825de8fa
  29. Huang J, Maram J, Tepelus TC, et al. Comparison of manual & automated analysis methods for corneal endothelial cell density measurements by specular microscopy. Journal of Optometry. 2018, 11(3): 182-191. https://doi.org/10.1016/j.optom.2017.06.001
  30. Maruoka S, Nakakura S, Matsuo N, et al. Comparison of semi-automated center-dot and fully automated endothelial cell analyses from specular microscopy images. International Ophthalmology. 2017, 38(6): 2495-2507. https://doi.org/10.1007/s10792-017-0760-7
  31. Heinzelmann S, Maier PC, Sandu K, et al. Assessing corneal endothelial cell density: automated versus manual countinJournal of Cataract and Refractive Surgery‌. 2016, 42: 1118–1125.
  32. Bourne WM. Cell Density in Corneal Endothelium: Effect of Age and Measurement Method. Investigative Ophthalmology & Visual Science‌. 1982, 2: 104-108.
  33. Laing RA, Sandstrom MM, Berrospi AR, et al. Changes in the corneal endothelium as a function of age. Experimental Eye Research‌. 1979, 29(4): 521-527.
  34. Brown NAP, Bron AJ. Endothelial cell size, shape, and apical structure in the aging human cornea. British Journal of Ophthalmology‌. 1977, 61(11): 805-814.
  35. McCarey BE, Edelhauser HF. The effects of age on the corneal endothelium. Ophthalmology. 1981, 88(6): 649-660.
  36. Bartels SP, Sperling M, Laing RA. The effect of specular microscope illumination on corneal endothelial cell count. Archives of Ophthalmology. 1981, 100(10): 1623-1626.
  37. Matsuda M, Yee RW, Edelhauser HF. Comparison of manual and automated endothelial cell counts in keratoconus. Ophthalmology. 1985, 92(9): 1146-1150.
  38. Waring GO, Bourne WM, Edelhauser HF, et al. The Corneal Endothelium. Ophthalmology. 1982, 89(6): 531-590. https://doi.org/10.1016/s0161-6420(82)34746-6
  39. Bourne WM, Nelson LR, Hodge DO. Cquot; Central corneal endothelial cell changes over ten years. Investigative Ophthalmology Camp; Visual Science. 1994, 35(7): 3023-3027.
  40. Laing RA, Sandstrom MM, Berrospi AR, et al. Changes in the corneal endothelium as a function of age. Ophthalmology. 1981, 88(4): 408-412.
  41. Hodges RR, Dartt DA, Jester JV. Cquot; Mechanisms of corneal endothelial cell migration. Ophthalmology. 1992, 99(3): 379-385.
  42. González P, Epstein DL. (1999). Endothelial cell behaviour after corneal surgery. Ophthalmic Surgery and Lasers. 1999, 30(5): 427-437.
  43. Møller-Pedersen T, Cavanagh HD, Petroll WM, et al. (1997). Cquot; Stromal wound healing explains refractive instability and haze development after photorefractive keratectomy: a 1-year confocal microscopic study. Ophthalmology. 1997, 104(10): 1564-1574.