Deep Learning Applied to Automated Segmentation of Geographic Atrophy in Fundus Autofluorescence Images

PURPOSE: This study describes the development of a deep learning algorithm based on the U-Net architecture for automated segmentation of geographic atrophy (GA) lesions in fundus autofluorescence (FAF) images. METHODS: Image preprocessing and normalization by modified adaptive histogram equalization...

Descripción completa

Detalles Bibliográficos
Autores principales: Arslan, Janan, Samarasinghe, Gihan, Sowmya, Arcot, Benke, Kurt K., Hodgson, Lauren A. B., Guymer, Robyn H., Baird, Paul N.
Formato: Online Artículo Texto
Lenguaje:English
Publicado: The Association for Research in Vision and Ophthalmology 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8267211/
https://www.ncbi.nlm.nih.gov/pubmed/34228106
http://dx.doi.org/10.1167/tvst.10.8.2
Descripción
Sumario:PURPOSE: This study describes the development of a deep learning algorithm based on the U-Net architecture for automated segmentation of geographic atrophy (GA) lesions in fundus autofluorescence (FAF) images. METHODS: Image preprocessing and normalization by modified adaptive histogram equalization were used for image standardization to improve effectiveness of deep learning. A U-Net–based deep learning algorithm was developed and trained and tested by fivefold cross-validation using FAF images from clinical datasets. The following metrics were used for evaluating the performance for lesion segmentation in GA: dice similarity coefficient (DSC), DSC loss, sensitivity, specificity, mean absolute error (MAE), accuracy, recall, and precision. RESULTS: In total, 702 FAF images from 51 patients were analyzed. After fivefold cross-validation for lesion segmentation, the average training and validation scores were found for the most important metric, DSC (0.9874 and 0.9779), for accuracy (0.9912 and 0.9815), for sensitivity (0.9955 and 0.9928), and for specificity (0.8686 and 0.7261). Scores for testing were all similar to the validation scores. The algorithm segmented GA lesions six times more quickly than human performance. CONCLUSIONS: The deep learning algorithm can be implemented using clinical data with a very high level of performance for lesion segmentation. Automation of diagnostics for GA assessment has the potential to provide savings with respect to patient visit duration, operational cost and measurement reliability in routine GA assessments. TRANSLATIONAL RELEVANCE: A deep learning algorithm based on the U-Net architecture and image preprocessing appears to be suitable for automated segmentation of GA lesions on clinical data, producing fast and accurate results.