Automated Detection of Central Retinal Artery Occlusion Using OCT Imaging via Explainable Deep Learning

Objective: To demonstrate the capability of a deep learning model to detect central retinal artery occlusion (CRAO), a retinal pathology with significant clinical urgency, using OCT data. Design: Retrospective, external validation study analyzing OCT and clinical baseline data of 2 institutions via...

Full description

Saved in:
Bibliographic Details
Main Authors: Ansgar Beuse, Daniel Alexander Wenzel, MD, Martin Stephan Spitzer, MD, Karl Ulrich Bartz-Schmidt, MD, Maximilian Schultheiss, MD, Sven Poli, MD, Carsten Grohmann, PhD, MD
Format: Article
Language:English
Published: Elsevier 2025-03-01
Series:Ophthalmology Science
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2666914524001660
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850220747217174528
author Ansgar Beuse
Daniel Alexander Wenzel, MD
Martin Stephan Spitzer, MD
Karl Ulrich Bartz-Schmidt, MD
Maximilian Schultheiss, MD
Sven Poli, MD
Carsten Grohmann, PhD, MD
author_facet Ansgar Beuse
Daniel Alexander Wenzel, MD
Martin Stephan Spitzer, MD
Karl Ulrich Bartz-Schmidt, MD
Maximilian Schultheiss, MD
Sven Poli, MD
Carsten Grohmann, PhD, MD
author_sort Ansgar Beuse
collection DOAJ
description Objective: To demonstrate the capability of a deep learning model to detect central retinal artery occlusion (CRAO), a retinal pathology with significant clinical urgency, using OCT data. Design: Retrospective, external validation study analyzing OCT and clinical baseline data of 2 institutions via deep learning classification analysis. Subjects: Patients presenting to the University Medical Center Tübingen and the University Medical Center Hamburg-Eppendorf in Germany. Methods: OCT data of patients suffering from CRAO, differential diagnosis with (sub) acute visual loss (central retinal vein occlusion, diabetic macular edema, nonarteritic ischemic optic neuropathy), and from controls were expertly graded and distinguished into 3 groups. Our methodological approach involved a nested multiclass five fold cross-validation classification scheme. Main Outcome Measures: Area under the curve (AUC). Results: The optimal performance of our algorithm was observed using 30 epochs, complemented by an early stopping mechanism to prevent overfitting. Our model followed a multiclass approach, distinguishing among the 3 different classes: control, CRAO, and differential diagnoses. The evaluation was conducted by the “one vs. all” area under the receiver operating characteristics curve (AUC) method. The results demonstrated AUC of 0.96 (95% confidence interval [CI], ± 0.01); 0.99 (95% CI, ± 0.00); and 0.90 (95% CI, ± 0.03) for each class, respectively. Conclusions: Our machine learning algorithm (MLA) exhibited a high AUC, as well as sensitivity and specificity in detecting CRAO and the differential classes, respectively. These findings underscore the potential for deploying MLAs in the identification of less common etiologies within an acute emergency clinical setting. Financial Disclosure(s): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.
format Article
id doaj-art-3f8fd1c5b3bc4094acb2fa8d1ddbb4c7
institution OA Journals
issn 2666-9145
language English
publishDate 2025-03-01
publisher Elsevier
record_format Article
series Ophthalmology Science
spelling doaj-art-3f8fd1c5b3bc4094acb2fa8d1ddbb4c72025-08-20T02:06:57ZengElsevierOphthalmology Science2666-91452025-03-015210063010.1016/j.xops.2024.100630Automated Detection of Central Retinal Artery Occlusion Using OCT Imaging via Explainable Deep LearningAnsgar Beuse0Daniel Alexander Wenzel, MD1Martin Stephan Spitzer, MD2Karl Ulrich Bartz-Schmidt, MD3Maximilian Schultheiss, MD4Sven Poli, MD5Carsten Grohmann, PhD, MD6Department of Ophthalmology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; Correspondence: Ansgar Beuse, Department of Ophthalmology, University Medical Center Hamburg-Eppendorf, Hamburg, D-20251, Germany.University Eye Hospital, Centre for Ophthalmology, University Hospital Tübingen, Tübingen, GermanyDepartment of Ophthalmology, University Medical Center Hamburg-Eppendorf, Hamburg, GermanyUniversity Eye Hospital, Centre for Ophthalmology, University Hospital Tübingen, Tübingen, GermanyDepartment of Ophthalmology, University Medical Center Hamburg-Eppendorf, Hamburg, GermanyDepartment of Neurology and Stroke, University Hospital Tübingen, Tübingen, Germany; Hertie Institute for Clinical Brain Research, University Hospital Tübingen, Tübingen, GermanyDepartment of Ophthalmology, University Medical Center Hamburg-Eppendorf, Hamburg, GermanyObjective: To demonstrate the capability of a deep learning model to detect central retinal artery occlusion (CRAO), a retinal pathology with significant clinical urgency, using OCT data. Design: Retrospective, external validation study analyzing OCT and clinical baseline data of 2 institutions via deep learning classification analysis. Subjects: Patients presenting to the University Medical Center Tübingen and the University Medical Center Hamburg-Eppendorf in Germany. Methods: OCT data of patients suffering from CRAO, differential diagnosis with (sub) acute visual loss (central retinal vein occlusion, diabetic macular edema, nonarteritic ischemic optic neuropathy), and from controls were expertly graded and distinguished into 3 groups. Our methodological approach involved a nested multiclass five fold cross-validation classification scheme. Main Outcome Measures: Area under the curve (AUC). Results: The optimal performance of our algorithm was observed using 30 epochs, complemented by an early stopping mechanism to prevent overfitting. Our model followed a multiclass approach, distinguishing among the 3 different classes: control, CRAO, and differential diagnoses. The evaluation was conducted by the “one vs. all” area under the receiver operating characteristics curve (AUC) method. The results demonstrated AUC of 0.96 (95% confidence interval [CI], ± 0.01); 0.99 (95% CI, ± 0.00); and 0.90 (95% CI, ± 0.03) for each class, respectively. Conclusions: Our machine learning algorithm (MLA) exhibited a high AUC, as well as sensitivity and specificity in detecting CRAO and the differential classes, respectively. These findings underscore the potential for deploying MLAs in the identification of less common etiologies within an acute emergency clinical setting. Financial Disclosure(s): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.http://www.sciencedirect.com/science/article/pii/S2666914524001660AI OCTCRAODeep learning retinaOCT imagingOphthalmology deep learning
spellingShingle Ansgar Beuse
Daniel Alexander Wenzel, MD
Martin Stephan Spitzer, MD
Karl Ulrich Bartz-Schmidt, MD
Maximilian Schultheiss, MD
Sven Poli, MD
Carsten Grohmann, PhD, MD
Automated Detection of Central Retinal Artery Occlusion Using OCT Imaging via Explainable Deep Learning
Ophthalmology Science
AI OCT
CRAO
Deep learning retina
OCT imaging
Ophthalmology deep learning
title Automated Detection of Central Retinal Artery Occlusion Using OCT Imaging via Explainable Deep Learning
title_full Automated Detection of Central Retinal Artery Occlusion Using OCT Imaging via Explainable Deep Learning
title_fullStr Automated Detection of Central Retinal Artery Occlusion Using OCT Imaging via Explainable Deep Learning
title_full_unstemmed Automated Detection of Central Retinal Artery Occlusion Using OCT Imaging via Explainable Deep Learning
title_short Automated Detection of Central Retinal Artery Occlusion Using OCT Imaging via Explainable Deep Learning
title_sort automated detection of central retinal artery occlusion using oct imaging via explainable deep learning
topic AI OCT
CRAO
Deep learning retina
OCT imaging
Ophthalmology deep learning
url http://www.sciencedirect.com/science/article/pii/S2666914524001660
work_keys_str_mv AT ansgarbeuse automateddetectionofcentralretinalarteryocclusionusingoctimagingviaexplainabledeeplearning
AT danielalexanderwenzelmd automateddetectionofcentralretinalarteryocclusionusingoctimagingviaexplainabledeeplearning
AT martinstephanspitzermd automateddetectionofcentralretinalarteryocclusionusingoctimagingviaexplainabledeeplearning
AT karlulrichbartzschmidtmd automateddetectionofcentralretinalarteryocclusionusingoctimagingviaexplainabledeeplearning
AT maximilianschultheissmd automateddetectionofcentralretinalarteryocclusionusingoctimagingviaexplainabledeeplearning
AT svenpolimd automateddetectionofcentralretinalarteryocclusionusingoctimagingviaexplainabledeeplearning
AT carstengrohmannphdmd automateddetectionofcentralretinalarteryocclusionusingoctimagingviaexplainabledeeplearning