Convolutional neural network to identify symptomatic Alzheimer's disease using multimodal retinal imaging.
Date
2020-11-26
Journal Title
Journal ISSN
Volume Title
Repository Usage Stats
views
downloads
Citation Stats
Attention Stats
Abstract
BACKGROUND/AIMS:To develop a convolutional neural network (CNN) to detect symptomatic Alzheimer's disease (AD) using a combination of multimodal retinal images and patient data. METHODS:Colour maps of ganglion cell-inner plexiform layer (GC-IPL) thickness, superficial capillary plexus (SCP) optical coherence tomography angiography (OCTA) images, and ultra-widefield (UWF) colour and fundus autofluorescence (FAF) scanning laser ophthalmoscopy images were captured in individuals with AD or healthy cognition. A CNN to predict AD diagnosis was developed using multimodal retinal images, OCT and OCTA quantitative data, and patient data. RESULTS:284 eyes of 159 subjects (222 eyes from 123 cognitively healthy subjects and 62 eyes from 36 subjects with AD) were used to develop the model. Area under the receiving operating characteristic curve (AUC) values for predicted probability of AD for the independent test set varied by input used: UWF colour AUC 0.450 (95% CI 0.282, 0.592), OCTA SCP 0.582 (95% CI 0.440, 0.724), UWF FAF 0.618 (95% CI 0.462, 0.773), GC-IPL maps 0.809 (95% CI 0.700, 0.919). A model incorporating all images, quantitative data and patient data (AUC 0.836 (CI 0.729, 0.943)) performed similarly to models only incorporating all images (AUC 0.829 (95% CI 0.719, 0.939)). GC-IPL maps, quantitative data and patient data AUC 0.841 (95% CI 0.739, 0.943). CONCLUSION:Our CNN used multimodal retinal images to successfully predict diagnosis of symptomatic AD in an independent test set. GC-IPL maps were the most useful single inputs for prediction. Models including only images performed similarly to models also including quantitative data and patient data.
Type
Department
Description
Provenance
Citation
Permalink
Published Version (Please cite this version)
Publication Info
Wisely, C Ellis, Dong Wang, Ricardo Henao, Dilraj S Grewal, Atalie C Thompson, Cason B Robbins, Stephen P Yoon, Srinath Soundararajan, et al. (2020). Convolutional neural network to identify symptomatic Alzheimer's disease using multimodal retinal imaging. The British journal of ophthalmology. 10.1136/bjophthalmol-2020-317659 Retrieved from https://hdl.handle.net/10161/21874.
This is constructed from limited available data and may be imprecise. To cite this article, please review & use the official citation provided by the journal.
Collections
Scholars@Duke
Ricardo Henao
Dilraj Singh Grewal
Vitreoretinal and Uveitis Specialist
Dilraj Grewal, MD specializes in the medical and surgical management of patients with complex Vitreoretinal pathology and Uveitis. He joined the Duke Eye Center in December 2016 following completion of his Vitreoretinal Surgery fellowship at Duke and Uveitis fellowship training at Moorfields Eye Hospital in London. Dr. Grewal is excited about treating patients with several of the new diagnostic and therapeutic modalities available as well as several others in the pipeline to better help patients with these potentially blinding diseases.
He has been the recipient of numerous prestigious awards including the Ronald G. Michels Foundation Fellowship Award, the Heed Ophthalmic Foundation Fellowship Award, Senior Achievement Award from the American Academy of Ophthalmology and Rhett Buckler and Senior Honor Awards from American Society of Retina Specialists.
Dr. Grewal has authored over 100 publications in peer-reviewed medical journals and over 150 presentations at national and international meetings. His research interests span clinical research activities in advanced ocular imaging and clinical trials for both Retina and Uveitis. He also serves as Director of Grading at the Duke Reading Center, a comprehensive image reading center that specializes in systematic analysis of ophthalmic images captured by many different modalities in multicenter clinical trials. In addition, he participates in national and international clinical trials in retina and uveitis.
Andrew John Liu
While striving to provide excellent clinical care, I also have several research interests:
1. Investigate a neurodevelopmental disorder, Tuberous Sclerosis Complex, which has the potential to provide insight into the pathophysiological mechanism of a neurodegenerative disorders, Alzheimer's Disease.
2. In collaboration with Dr. Cathrine Hoyo, we are investigating an epigenetic mechanism to explain the racial disparities in the development of Alzheimer's disease between underrepresented minorities and European Americans.
3. I am interested in clinically characterizing the long-term cognitive and behavioral consequences in convalescent COVID-19 patients.
Sharon Fekrat
Sharon Fekrat, MD, FACS, FASRS is a board-certified, fellowship-trained vitreoretinal surgeon and an accomplished educator, clinical researcher, and administrator. She was the first woman Assistant Chief of Service who was on-call handling ocular trauma and retinal emergencies continuously for 365 days at Johns Hopkins Wilmer Eye Institute, one of the top institutions in the world, before she was recruited to Duke. Dr. Fekrat's achievements in patient care and vitreoretinal surgery have been recognized by her peers in "best doctors" lists, including in Business North Carolina since 2005, Newsweek, Ocular Surgery News, as well as her induction into the Retina Hall of Fame. Dr. Fekrat was awarded the Secretariat Award and Achievement Award from the American Academy of Ophthalmology, Rhett Buckler Trophies for surgical video competition and Senior Honor Award from the American Society of Retina Specialists, the Janet M. Glasgow Memorial Achievement Award from the American Medical Women's Association, Ronald G. Michels Fellowship Foundation Award (first woman), and Heed and Heed-Knapp Awards.
Dr. Fekrat is an inspirational and committed teacher, mentor, and sponsor of future ophthalmologists and retina specialists whom she trains. She has received the Faculty Instructor of the Year from Duke Ophthalmic Medical Technician Training Program, Golden Globe Award from the Duke Ophthalmology Residents, and the 2022 ASRS Crystal Apple Award from the Young Career Section of the American Society of Retina Specialists. She taught vitreoretinal surgery fellows as invited faculty at the Fort Worth Vitreoretinal Surgery Training Course and Wetlab and the Duke Fellows Advanced Vitreous Surgery Course and Wetlab. Dr. Fekrat also mentors numerous undergraduates, medical students, residents, and fellows in clinical research, publishing from her retinal vein occlusion, endophthalmitis, vitreomacular traction, keratoprosthesis, and submacular hemorrhage databases.
Dr. Fekrat founded and leads the iMIND international multidisciplinary clinical research team evaluating multimodal retinal, choroidal, and optic nerve imaging for the diagnosis of Alzheimer's disease and other neurodegenerations and is collaborating with Duke engineers and computer scientists to train machine learning models for the diagnosis of Alzheimer's, Parkinson's, and mild cognitive impairment as well as image quality assessment. Her team published the first paper demonstrating proof of concept that a machine learning model can differentiate individuals with Alzheimer's disease from those with normal cognition using retinal images. She is also studying Alzheimer's proteins in the aqueous humor of the eye. Her team's work has been featured on ABC, CBS, Fox, Reuters, Newsweek, People's Pharmacy, and several hundred news outlets around the world. Her research work has been recognized with numerous awards including the VitreoRetinal Surgery Foundation Award, Karen L Wrenn Alzheimer's Disease Travel Award, Robert Machemer Resident Research Award, American Society of Retina Specialists Journal of VitreoRetinal Diseases Distinguished Contributor Award, Research to Prevent Blindness Departmental Small Grant Award, Research to Prevent Blindness Medical Student Eye Research Fellowship, Vit-Buckle Society BullDogger and Academic Grant Awards, Women in Retina Travel Grant, Gerhard Zinser Memorial Travel Grant, Alzheimer's Disease Drug Discovery Foundation, and many more. As an invited speaker, she was keynote speaker at Research to Prevent Blindness' Vision Research Funding Partnership speaking on Artificial Intelligence and the Retina: Potential for Diagnosis, Challenges for Progress. She also gave invited testimony on her iMIND work to the United States Senate Special Committee on Aging.
Dr. Fekrat has published extensively not only in medical journals but also in textbooks. She has co-authored over 200 publications in peer-reviewed medical journals and over 50 textbook chapters. She is chief Editor of Duke Eye Center’s All About Your Eyes for the lay public, Wolters Kluwer's The Duke Manual of Vitreoretinal Surgery and both editions of SLACK's Curbside Consultation in Retina as well as Series Editor for The Duke Manuals of Ophthalmic Surgery for ophthalmic microsurgeons all over the world. Dr Fekrat is Faculty Editor-in-Chief of the Duke Journal of Case Reports in Ophthalmology which she founded. Several scientific and scholarly journals have sought Dr. Fekrat's expertise and recognized her research leadership by naming her to their editorial boards and seeking her assistance in reviewing the work of other investigators and scientists. She serves on the editorial board of Retinal Physician, Ophthalmology Times, Ophthalmic Surgery Lasers Imaging Retina, and American Society of Retina Specialists' Retina Times, and is an Executive Editor of the American Journal of Ophthalmology. She has also served as Retina Section Editor and Ophthalmic Pearls section Co-Editor of EyeNet, published by the American Academy of Ophthalmology.
Dr. Fekrat is a natural leader and administrator evident in the breadth and depth of her leadership roles. To mention a few, she has been Director of Duke Medical Student Ophthalmology Education, Director of the Duke Vitreoretinal Surgery Fellowship Program (one of the best in the world), Director of Ophthalmology Faculty Mentoring and Career Development, President of the North Carolina Society of Eye Physicians and Surgeons, Vice Chair of Duke Clinical Sciences Appointments, Promotions, and Tenure Committee, and is currently Vice Chair of Faculty Affairs and Director of Duke iMIND Research Group. She was appointed as the sole physician on the Steering Committee for Strategic Education on Research, Translation, and Commercialization that advised the Duke Board of Trustees. At Duke's VA affiliate (Durham VA Health Care System), she has served as Chief of Ophthalmology, Interim Chief of Surgery (leading 80 Duke surgeons across all surgical specialties), Interim Deputy Chief of Staff, and is currently Associate Chief of Staff of Microsurgery and Specialty Services.
Dr. Fekrat has an impressive educational pedigree. After graduating as valedictorian of her high school, she received her Bachelors of Science degree from Georgetown University graduating magna cum laude and Phi Beta Kappa and spent one year at Oxford University in England. She received her MD from the University of Chicago Pritzker School of Medicine, recognized by Alpha Omega Alpha Honor Society. She completed her ophthalmology residency training at the prestigious Johns Hopkins Wilmer Eye Institute where she also completed her two year fellowship in medical and surgical vitreoretinal diseases with the giants in the field of ophthalmology and retina.
https://dukeeyecenter.duke.edu/research/clinical-research/clinical-research-programs/imind
https://dukeeyecenter.duke.edu/djcro
https://www.linkedin.com/in/sharon-fekrat-md-facs-fasrs-315a31a6/
Unless otherwise indicated, scholarly articles published by Duke faculty members are made available here with a CC-BY-NC (Creative Commons Attribution Non-Commercial) license, as enabled by the Duke Open Access Policy. If you wish to use the materials in ways not already permitted under CC-BY-NC, please consult the copyright owner. Other materials are made available here through the author’s grant of a non-exclusive license to make their work openly accessible.