Open AccessVisualising ganglion cell layer based on image entropy optimisation for adaptive contrast enhancement
Author(s) -
Han J.H.,
Cha J.
Publication year - 2020
Publication title -
electronics letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.375
H-Index - 146
ISSN - 1350-911X
DOI - 10.1049/el.2019.3006
Subject(s) - artificial intelligence , computer vision , segmentation , ganglion cell layer , optical coherence tomography , computer science , image segmentation , nerve fiber layer , pattern recognition (psychology) , image contrast , contrast (vision) , entropy (arrow of time) , glaucoma , retinal , optics , ophthalmology , medicine , physics , quantum mechanics
Optical coherence tomography cannot easily be used for visual identification of the ganglion cell layer (GCL) for diagnosing retinal diseases owing to the extremely low image contrast between adjacent layers. To solve this problem, the authors used a limit‐clipping optimisation method along with the image entropy to enhance the image contrast of targeted layers. As a result, the GCL was successfully extracted using an intelligent tracking system without impacting the segmentation of other retinal layers and image morphology. The segmentation results were evaluated through comparisons with manual segmentation results provided by clinical experts. The results of this study should help realise simple and efficient discrimination of important retinal layers for the early diagnosis of glaucoma.