Internal validation of a convolutional neural network pipeline for assessing meibomian gland structure from meibography
Comparison of predictions from each model to expert panel ground-truth demonstrated strong association and moderate to substantial agreement. The findings and performance metrics show that the pipeline of algorithms provides standardized, real-time inference/prediction of meibomian gland absence.
Authors
- Caroline A. Blackie (ORCID: https://orcid.org/0000-0002-4618-3303)
- Preeya K. Gupta (ORCID: https://orcid.org/0000-0003-0895-7870)
- Daniel Cheng (ORCID: https://orcid.org/0000-0002-5247-7113)
- Edward J. Holland (ORCID: https://orcid.org/0000-0001-7732-3172)
- Joshua Young
- Charles Scales (ORCID: https://orcid.org/0009-0007-7512-3944)
- David Kading
- Whitney Hauser
- John Y. H. Bai (ORCID: https://orcid.org/0000-0001-6688-7837)
- Eugene Appenteng Osae (ORCID: https://orcid.org/0000-0002-2059-7823)
- David Murakami (ORCID: https://orcid.org/0009-0006-1939-1947)
- Leslie O’Dell (ORCID: https://orcid.org/0000-0003-3017-6800)
- Casey Claypool
Institutions
- Johnson & Johnson (United States) (US)
- Duke University (US)
- Mississippi Division of Medicaid (US)
- Merck Canada Inc. (Canada) (CA)
- American Academy of Optometry (US)
- Empire Eye Physicians (US)
- Cincinnati Eye Institute (US)
- LEK Consulting (United States) (US)
- Dompé (Italy) (IT)
- Johnson & Johnson (Canada) (CA)
- New York University (US)
- SUNY College of Optometry (US)
Publication Details
- Journal
- Optometry and Vision Science
- Published
- 2025-01-01
- DOI
- https://doi.org/10.1097/opx.0000000000002208
- Citations
- 1
- Primary Topic
- Ocular Surface and Contact Lens
- Type
- article
- Field-Weighted Citation Impact
- 1.88
Funders
- Johnson and Johnson Vision Care