Cross-instrument optical coherence tomography-angiography (OCTA)-based prediction of age-related macular degeneration (AMD) disease activity using artificial intelligence

Abstract This study investigates the efficacy of predicting age-related macular degeneration (AMD) activity through deep neural networks (DNN) using a cross-instrument training dataset composed of Optical coherence tomography-angiography (OCTA) images from two different manufacturers. A retrospectiv...

Full description

Saved in:
Bibliographic Details
Main Authors: Anna Heinke, Haochen Zhang, Krzysztof Broniarek, Katarzyna Michalska-Małecka, Wyatt Elsner, Carlo Miguel B. Galang, Daniel N. Deussen, Alexandra Warter, Fritz Kalaw, Ines Nagel, Akshay Agnihotri, Nehal N. Mehta, Julian Elias Klaas, Valerie Schmelter, Igor Kozak, Sally L. Baxter, Dirk-Uwe Bartsch, Lingyun Cheng, Cheolhong An, Truong Nguyen, William R. Freeman
Format: Article
Language:English
Published: Nature Portfolio 2024-11-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-024-78327-0
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846172070301401088
author Anna Heinke
Haochen Zhang
Krzysztof Broniarek
Katarzyna Michalska-Małecka
Wyatt Elsner
Carlo Miguel B. Galang
Daniel N. Deussen
Alexandra Warter
Fritz Kalaw
Ines Nagel
Akshay Agnihotri
Nehal N. Mehta
Julian Elias Klaas
Valerie Schmelter
Igor Kozak
Sally L. Baxter
Dirk-Uwe Bartsch
Lingyun Cheng
Cheolhong An
Truong Nguyen
William R. Freeman
author_facet Anna Heinke
Haochen Zhang
Krzysztof Broniarek
Katarzyna Michalska-Małecka
Wyatt Elsner
Carlo Miguel B. Galang
Daniel N. Deussen
Alexandra Warter
Fritz Kalaw
Ines Nagel
Akshay Agnihotri
Nehal N. Mehta
Julian Elias Klaas
Valerie Schmelter
Igor Kozak
Sally L. Baxter
Dirk-Uwe Bartsch
Lingyun Cheng
Cheolhong An
Truong Nguyen
William R. Freeman
author_sort Anna Heinke
collection DOAJ
description Abstract This study investigates the efficacy of predicting age-related macular degeneration (AMD) activity through deep neural networks (DNN) using a cross-instrument training dataset composed of Optical coherence tomography-angiography (OCTA) images from two different manufacturers. A retrospective cross-sectional study analyzed 2D vascular en-face OCTA images from Heidelberg Spectralis (1478 samples: 1102 training, 276 validation, 100 testing) and Optovue Solix (1003 samples: 754 training, 189 validation, 60 testing). OCTA scans were labeled based on clinical diagnoses and adjacent B-scan OCT fluid information, categorizing activity into normal, dry AMD, active wet AMD, and wet AMD in remission. Experiments explored cross-instrument disease classification using separate and combined datasets for training the DNN. Testing involved 100 Heidelberg and 60 Optovue samples. Training on Heidelberg data alone yielded 73% accuracy on Heidelberg images and 60% on Optovue images. Training on Optovue data alone resulted in 34% accuracy on Heidelberg and 85% on Optovue images. Combined training data from both instruments achieved 78% accuracy on Heidelberg and 76% on Optovue test sets. Results indicate that cross-instrument classifier training demonstrates high classification prediction accuracy, making cross-instrument training viable for future clinical applications. This implies that vascular morphology in OCTA can predict disease progression.
format Article
id doaj-art-f0b1c73441e648c8969d26e7c6921c1c
institution Kabale University
issn 2045-2322
language English
publishDate 2024-11-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-f0b1c73441e648c8969d26e7c6921c1c2024-11-10T12:22:30ZengNature PortfolioScientific Reports2045-23222024-11-011411910.1038/s41598-024-78327-0Cross-instrument optical coherence tomography-angiography (OCTA)-based prediction of age-related macular degeneration (AMD) disease activity using artificial intelligenceAnna Heinke0Haochen Zhang1Krzysztof Broniarek2Katarzyna Michalska-Małecka3Wyatt Elsner4Carlo Miguel B. Galang5Daniel N. Deussen6Alexandra Warter7Fritz Kalaw8Ines Nagel9Akshay Agnihotri10Nehal N. Mehta11Julian Elias Klaas12Valerie Schmelter13Igor Kozak14Sally L. Baxter15Dirk-Uwe Bartsch16Lingyun Cheng17Cheolhong An18Truong Nguyen19William R. Freeman20Jacobs Retina CenterDepartment of Electrical and Computer Engineering, University of California San DiegoDepartment of Ophthalmology, Medical University of GdańskDepartment of Ophthalmology, Medical University of GdańskThe Department of Cognitive Science, University of California San DiegoJacobs Retina CenterJacobs Retina CenterJacobs Retina CenterJacobs Retina CenterJacobs Retina CenterJacobs Retina CenterJacobs Retina CenterDepartment of Ophthalmology, LMU University Hospital, LMU MunichDepartment of Ophthalmology, LMU University Hospital, LMU MunichMoorfields Eye HospitalViterbi Family Department of Ophthalmology and Shiley Eye Institute, University of California San DiegoJacobs Retina CenterJacobs Retina CenterDepartment of Electrical and Computer Engineering, University of California San DiegoDepartment of Electrical and Computer Engineering, University of California San DiegoJacobs Retina CenterAbstract This study investigates the efficacy of predicting age-related macular degeneration (AMD) activity through deep neural networks (DNN) using a cross-instrument training dataset composed of Optical coherence tomography-angiography (OCTA) images from two different manufacturers. A retrospective cross-sectional study analyzed 2D vascular en-face OCTA images from Heidelberg Spectralis (1478 samples: 1102 training, 276 validation, 100 testing) and Optovue Solix (1003 samples: 754 training, 189 validation, 60 testing). OCTA scans were labeled based on clinical diagnoses and adjacent B-scan OCT fluid information, categorizing activity into normal, dry AMD, active wet AMD, and wet AMD in remission. Experiments explored cross-instrument disease classification using separate and combined datasets for training the DNN. Testing involved 100 Heidelberg and 60 Optovue samples. Training on Heidelberg data alone yielded 73% accuracy on Heidelberg images and 60% on Optovue images. Training on Optovue data alone resulted in 34% accuracy on Heidelberg and 85% on Optovue images. Combined training data from both instruments achieved 78% accuracy on Heidelberg and 76% on Optovue test sets. Results indicate that cross-instrument classifier training demonstrates high classification prediction accuracy, making cross-instrument training viable for future clinical applications. This implies that vascular morphology in OCTA can predict disease progression.https://doi.org/10.1038/s41598-024-78327-0Optical coherence tomography-angiography (OCTA)Cross-instrument trainingArtificial intelligence (AI)Deep neural networks (DNN)Age-related macular degeneration (AMD)
spellingShingle Anna Heinke
Haochen Zhang
Krzysztof Broniarek
Katarzyna Michalska-Małecka
Wyatt Elsner
Carlo Miguel B. Galang
Daniel N. Deussen
Alexandra Warter
Fritz Kalaw
Ines Nagel
Akshay Agnihotri
Nehal N. Mehta
Julian Elias Klaas
Valerie Schmelter
Igor Kozak
Sally L. Baxter
Dirk-Uwe Bartsch
Lingyun Cheng
Cheolhong An
Truong Nguyen
William R. Freeman
Cross-instrument optical coherence tomography-angiography (OCTA)-based prediction of age-related macular degeneration (AMD) disease activity using artificial intelligence
Scientific Reports
Optical coherence tomography-angiography (OCTA)
Cross-instrument training
Artificial intelligence (AI)
Deep neural networks (DNN)
Age-related macular degeneration (AMD)
title Cross-instrument optical coherence tomography-angiography (OCTA)-based prediction of age-related macular degeneration (AMD) disease activity using artificial intelligence
title_full Cross-instrument optical coherence tomography-angiography (OCTA)-based prediction of age-related macular degeneration (AMD) disease activity using artificial intelligence
title_fullStr Cross-instrument optical coherence tomography-angiography (OCTA)-based prediction of age-related macular degeneration (AMD) disease activity using artificial intelligence
title_full_unstemmed Cross-instrument optical coherence tomography-angiography (OCTA)-based prediction of age-related macular degeneration (AMD) disease activity using artificial intelligence
title_short Cross-instrument optical coherence tomography-angiography (OCTA)-based prediction of age-related macular degeneration (AMD) disease activity using artificial intelligence
title_sort cross instrument optical coherence tomography angiography octa based prediction of age related macular degeneration amd disease activity using artificial intelligence
topic Optical coherence tomography-angiography (OCTA)
Cross-instrument training
Artificial intelligence (AI)
Deep neural networks (DNN)
Age-related macular degeneration (AMD)
url https://doi.org/10.1038/s41598-024-78327-0
work_keys_str_mv AT annaheinke crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT haochenzhang crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT krzysztofbroniarek crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT katarzynamichalskamałecka crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT wyattelsner crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT carlomiguelbgalang crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT danielndeussen crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT alexandrawarter crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT fritzkalaw crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT inesnagel crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT akshayagnihotri crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT nehalnmehta crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT julianeliasklaas crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT valerieschmelter crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT igorkozak crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT sallylbaxter crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT dirkuwebartsch crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT lingyuncheng crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT cheolhongan crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT truongnguyen crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence
AT williamrfreeman crossinstrumentopticalcoherencetomographyangiographyoctabasedpredictionofagerelatedmaculardegenerationamddiseaseactivityusingartificialintelligence