Deep learning-based automatic meibomian gland segmentation and morphology assessment in infrared meibography
Abstract Meibomian glands (MG) are large sebaceous glands located below the tarsal conjunctiva and the abnormalities of these glands cause Meibomian gland dysfunction (MGD) which is responsible for evaporative dry eye disease (DED). Accurate MG segmentation is a key prerequisite for automated imagin...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2021-04-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-021-87314-8 |
| _version_ | 1818834935678500864 |
|---|---|
| author | Md Asif Khan Setu Jens Horstmann Stefan Schmidt Michael E. Stern Philipp Steven |
| author_facet | Md Asif Khan Setu Jens Horstmann Stefan Schmidt Michael E. Stern Philipp Steven |
| author_sort | Md Asif Khan Setu |
| collection | DOAJ |
| description | Abstract Meibomian glands (MG) are large sebaceous glands located below the tarsal conjunctiva and the abnormalities of these glands cause Meibomian gland dysfunction (MGD) which is responsible for evaporative dry eye disease (DED). Accurate MG segmentation is a key prerequisite for automated imaging based MGD related DED diagnosis. However, Automatic MG segmentation in infrared meibography is a challenging task due to image artifacts. A deep learning-based MG segmentation has been proposed which directly learns MG features from the training image dataset without any image pre-processing. The model is trained and evaluated using 728 anonymized clinical meibography images. Additionally, automatic MG morphometric parameters, gland number, length, width, and tortuosity assessment were proposed. The average precision, recall, and F1 score were achieved 83%, 81%, and 84% respectively on the testing dataset with AUC value of 0.96 based on ROC curve and dice coefficient of 84%. Single image segmentation and morphometric parameter evaluation took on average 1.33 s. To the best of our knowledge, this is the first time that a validated deep learning-based approach is applied in MG segmentation and evaluation for both upper and lower eyelids. |
| first_indexed | 2024-12-19T02:42:43Z |
| format | Article |
| id | doaj.art-81521c54131d43678c2babeb7875e399 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-12-19T02:42:43Z |
| publishDate | 2021-04-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-81521c54131d43678c2babeb7875e3992022-12-21T20:39:06ZengNature PortfolioScientific Reports2045-23222021-04-0111111110.1038/s41598-021-87314-8Deep learning-based automatic meibomian gland segmentation and morphology assessment in infrared meibographyMd Asif Khan Setu0Jens Horstmann1Stefan Schmidt2Michael E. Stern3Philipp Steven4Department of Ophthalmology, Faculty of Medicine, University Hospital Cologne, University of CologneDepartment of Ophthalmology, Faculty of Medicine, University Hospital Cologne, University of CologneHeidelberg Engineering GmbHDepartment of Ophthalmology, Faculty of Medicine, University Hospital Cologne, University of CologneDepartment of Ophthalmology, Faculty of Medicine, University Hospital Cologne, University of CologneAbstract Meibomian glands (MG) are large sebaceous glands located below the tarsal conjunctiva and the abnormalities of these glands cause Meibomian gland dysfunction (MGD) which is responsible for evaporative dry eye disease (DED). Accurate MG segmentation is a key prerequisite for automated imaging based MGD related DED diagnosis. However, Automatic MG segmentation in infrared meibography is a challenging task due to image artifacts. A deep learning-based MG segmentation has been proposed which directly learns MG features from the training image dataset without any image pre-processing. The model is trained and evaluated using 728 anonymized clinical meibography images. Additionally, automatic MG morphometric parameters, gland number, length, width, and tortuosity assessment were proposed. The average precision, recall, and F1 score were achieved 83%, 81%, and 84% respectively on the testing dataset with AUC value of 0.96 based on ROC curve and dice coefficient of 84%. Single image segmentation and morphometric parameter evaluation took on average 1.33 s. To the best of our knowledge, this is the first time that a validated deep learning-based approach is applied in MG segmentation and evaluation for both upper and lower eyelids.https://doi.org/10.1038/s41598-021-87314-8 |
| spellingShingle | Md Asif Khan Setu Jens Horstmann Stefan Schmidt Michael E. Stern Philipp Steven Deep learning-based automatic meibomian gland segmentation and morphology assessment in infrared meibography Scientific Reports |
| title | Deep learning-based automatic meibomian gland segmentation and morphology assessment in infrared meibography |
| title_full | Deep learning-based automatic meibomian gland segmentation and morphology assessment in infrared meibography |
| title_fullStr | Deep learning-based automatic meibomian gland segmentation and morphology assessment in infrared meibography |
| title_full_unstemmed | Deep learning-based automatic meibomian gland segmentation and morphology assessment in infrared meibography |
| title_short | Deep learning-based automatic meibomian gland segmentation and morphology assessment in infrared meibography |
| title_sort | deep learning based automatic meibomian gland segmentation and morphology assessment in infrared meibography |
| url | https://doi.org/10.1038/s41598-021-87314-8 |
| work_keys_str_mv | AT mdasifkhansetu deeplearningbasedautomaticmeibomianglandsegmentationandmorphologyassessmentininfraredmeibography AT jenshorstmann deeplearningbasedautomaticmeibomianglandsegmentationandmorphologyassessmentininfraredmeibography AT stefanschmidt deeplearningbasedautomaticmeibomianglandsegmentationandmorphologyassessmentininfraredmeibography AT michaelestern deeplearningbasedautomaticmeibomianglandsegmentationandmorphologyassessmentininfraredmeibography AT philippsteven deeplearningbasedautomaticmeibomianglandsegmentationandmorphologyassessmentininfraredmeibography |