Your browser doesn't support javascript.
loading
Meibomian glands segmentation in infrared images with limited annotation.
Lin, Jia-Wen; Lin, Ling-Jie; Lu, Feng; Lai, Tai-Chen; Zou, Jing; Guo, Lin-Ling; Lin, Zhi-Ming; Li, Li.
Afiliación
  • Lin JW; College of Computer and Data Science, Fuzhou University, Fuzhou 350108, Fujian Province, China.
  • Lin LJ; Fujian Provincial Key Laboratory of Networking Computing and Intelligent Information Processing, Fuzhou University, Fuzhou 350108, Fujian Province, China.
  • Lu F; College of Computer and Data Science, Fuzhou University, Fuzhou 350108, Fujian Province, China.
  • Lai TC; Fujian Provincial Key Laboratory of Networking Computing and Intelligent Information Processing, Fuzhou University, Fuzhou 350108, Fujian Province, China.
  • Zou J; College of Computer and Data Science, Fuzhou University, Fuzhou 350108, Fujian Province, China.
  • Guo LL; Fujian Provincial Key Laboratory of Networking Computing and Intelligent Information Processing, Fuzhou University, Fuzhou 350108, Fujian Province, China.
  • Lin ZM; Shengli Clinical Medical College of Fujian Medical University, Fuzhou 350001, Fujian Province, China.
  • Li L; Shengli Clinical Medical College of Fujian Medical University, Fuzhou 350001, Fujian Province, China.
Int J Ophthalmol ; 17(3): 401-407, 2024.
Article en En | MEDLINE | ID: mdl-38721512
ABSTRACT

AIM:

To investigate a pioneering framework for the segmentation of meibomian glands (MGs), using limited annotations to reduce the workload on ophthalmologists and enhance the efficiency of clinical diagnosis.

METHODS:

Totally 203 infrared meibomian images from 138 patients with dry eye disease, accompanied by corresponding annotations, were gathered for the study. A rectified scribble-supervised gland segmentation (RSSGS) model, incorporating temporal ensemble prediction, uncertainty estimation, and a transformation equivariance constraint, was introduced to address constraints imposed by limited supervision information inherent in scribble annotations. The viability and efficacy of the proposed model were assessed based on accuracy, intersection over union (IoU), and dice coefficient.

RESULTS:

Using manual labels as the gold standard, RSSGS demonstrated outcomes with an accuracy of 93.54%, a dice coefficient of 78.02%, and an IoU of 64.18%. Notably, these performance metrics exceed the current weakly supervised state-of-the-art methods by 0.76%, 2.06%, and 2.69%, respectively. Furthermore, despite achieving a substantial 80% reduction in annotation costs, it only lags behind fully annotated methods by 0.72%, 1.51%, and 2.04%.

CONCLUSION:

An innovative automatic segmentation model is developed for MGs in infrared eyelid images, using scribble annotation for training. This model maintains an exceptionally high level of segmentation accuracy while substantially reducing training costs. It holds substantial utility for calculating clinical parameters, thereby greatly enhancing the diagnostic efficiency of ophthalmologists in evaluating meibomian gland dysfunction.
Palabras clave
Texto completo
Añadir a Mi BVS
Imprimir
XML
PubMed Links
Buscar en Google

Texto completo: 1 Colección: 01-internacional Base de datos: MEDLINE Idioma: En Revista: Int J Ophthalmol Año: 2024 Tipo del documento: Article País de afiliación: China Pais de publicación: China
Texto completo
Añadir a Mi BVS
Imprimir
XML
PubMed Links
Buscar en Google

Texto completo: 1 Colección: 01-internacional Base de datos: MEDLINE Idioma: En Revista: Int J Ophthalmol Año: 2024 Tipo del documento: Article País de afiliación: China Pais de publicación: China