Separate Detection of Stromal and Epithelial Corneal Edema on Optical Coherence Tomography Using a Deep Learning Pipeline and Transfer Learning
The accurate detection of corneal edema has become a topic of growing interest with the generalization of endothelial keratoplasty. Despite recent advances in deep learning for corneal edema detection, the problem of minimal edema remains challenging. Using transfer learning and a limited training s...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-10-01
|
Series: | Photonics |
Subjects: | |
Online Access: | https://www.mdpi.com/2304-6732/8/11/483 |
_version_ | 1797508745575530496 |
---|---|
author | Pierre Zéboulon Wassim Ghazal Karen Bitton Damien Gatinel |
author_facet | Pierre Zéboulon Wassim Ghazal Karen Bitton Damien Gatinel |
author_sort | Pierre Zéboulon |
collection | DOAJ |
description | The accurate detection of corneal edema has become a topic of growing interest with the generalization of endothelial keratoplasty. Despite recent advances in deep learning for corneal edema detection, the problem of minimal edema remains challenging. Using transfer learning and a limited training set of 11 images, we built a model to segment the corneal epithelium, which is part of a three-model pipeline to detect corneal edema. A second and a third model are used to detect edema on the stroma alone and on the epithelium. A validation set of 233 images from 30 patients consisting of three groups (Normal, Minimal Edema and important Edema) was used to compare the results of our new pipeline to our previous model. The mean edema fraction (EF), defined as the number of pixels detected as edema divided by the total number of pixels of the cornea, was calculated for each image. With our previous model, the mean EF was not statistically different between the Normal and Minimal Edema groups (<i>p</i> = 0.24). With the current pipeline, the mean EF was higher in the Minimal Edema group compared to the Normal group (<i>p</i> < 0.01). The described pipeline constitutes an adjustable framework for the detection of corneal edema based on optical coherence tomography and yields better performances in cases of minimal or localized edema. |
first_indexed | 2024-03-10T05:09:14Z |
format | Article |
id | doaj.art-a8a1cbadf0da496da69e383adacd2d0e |
institution | Directory Open Access Journal |
issn | 2304-6732 |
language | English |
last_indexed | 2024-03-10T05:09:14Z |
publishDate | 2021-10-01 |
publisher | MDPI AG |
record_format | Article |
series | Photonics |
spelling | doaj.art-a8a1cbadf0da496da69e383adacd2d0e2023-11-23T01:01:17ZengMDPI AGPhotonics2304-67322021-10-0181148310.3390/photonics8110483Separate Detection of Stromal and Epithelial Corneal Edema on Optical Coherence Tomography Using a Deep Learning Pipeline and Transfer LearningPierre Zéboulon0Wassim Ghazal1Karen Bitton2Damien Gatinel3Anterior and Refractive Surgery Department, Rothschild Foundation Hospital, 75019 Paris, FranceAnterior and Refractive Surgery Department, Rothschild Foundation Hospital, 75019 Paris, FranceAnterior and Refractive Surgery Department, Rothschild Foundation Hospital, 75019 Paris, FranceAnterior and Refractive Surgery Department, Rothschild Foundation Hospital, 75019 Paris, FranceThe accurate detection of corneal edema has become a topic of growing interest with the generalization of endothelial keratoplasty. Despite recent advances in deep learning for corneal edema detection, the problem of minimal edema remains challenging. Using transfer learning and a limited training set of 11 images, we built a model to segment the corneal epithelium, which is part of a three-model pipeline to detect corneal edema. A second and a third model are used to detect edema on the stroma alone and on the epithelium. A validation set of 233 images from 30 patients consisting of three groups (Normal, Minimal Edema and important Edema) was used to compare the results of our new pipeline to our previous model. The mean edema fraction (EF), defined as the number of pixels detected as edema divided by the total number of pixels of the cornea, was calculated for each image. With our previous model, the mean EF was not statistically different between the Normal and Minimal Edema groups (<i>p</i> = 0.24). With the current pipeline, the mean EF was higher in the Minimal Edema group compared to the Normal group (<i>p</i> < 0.01). The described pipeline constitutes an adjustable framework for the detection of corneal edema based on optical coherence tomography and yields better performances in cases of minimal or localized edema.https://www.mdpi.com/2304-6732/8/11/483deep learningcorneal edemaFuchs endothelial corneal dystrophyoptical coherence tomography |
spellingShingle | Pierre Zéboulon Wassim Ghazal Karen Bitton Damien Gatinel Separate Detection of Stromal and Epithelial Corneal Edema on Optical Coherence Tomography Using a Deep Learning Pipeline and Transfer Learning Photonics deep learning corneal edema Fuchs endothelial corneal dystrophy optical coherence tomography |
title | Separate Detection of Stromal and Epithelial Corneal Edema on Optical Coherence Tomography Using a Deep Learning Pipeline and Transfer Learning |
title_full | Separate Detection of Stromal and Epithelial Corneal Edema on Optical Coherence Tomography Using a Deep Learning Pipeline and Transfer Learning |
title_fullStr | Separate Detection of Stromal and Epithelial Corneal Edema on Optical Coherence Tomography Using a Deep Learning Pipeline and Transfer Learning |
title_full_unstemmed | Separate Detection of Stromal and Epithelial Corneal Edema on Optical Coherence Tomography Using a Deep Learning Pipeline and Transfer Learning |
title_short | Separate Detection of Stromal and Epithelial Corneal Edema on Optical Coherence Tomography Using a Deep Learning Pipeline and Transfer Learning |
title_sort | separate detection of stromal and epithelial corneal edema on optical coherence tomography using a deep learning pipeline and transfer learning |
topic | deep learning corneal edema Fuchs endothelial corneal dystrophy optical coherence tomography |
url | https://www.mdpi.com/2304-6732/8/11/483 |
work_keys_str_mv | AT pierrezeboulon separatedetectionofstromalandepithelialcornealedemaonopticalcoherencetomographyusingadeeplearningpipelineandtransferlearning AT wassimghazal separatedetectionofstromalandepithelialcornealedemaonopticalcoherencetomographyusingadeeplearningpipelineandtransferlearning AT karenbitton separatedetectionofstromalandepithelialcornealedemaonopticalcoherencetomographyusingadeeplearningpipelineandtransferlearning AT damiengatinel separatedetectionofstromalandepithelialcornealedemaonopticalcoherencetomographyusingadeeplearningpipelineandtransferlearning |