Posted on 15 May 2023 | The copyright holder is the author/funder. All rights reserved. No reuse without permission. | https://doi.org/10.22541/au.168412879.91914539/v1 | This a preprint and has not been peer reviewed. Data may be preliminary. Automated diagnosis of AD using OCT and OCTA: A systematic review Yasemin Turkan 1 and Faik Boray Tek 2 1 Isik University 2 Istanbul Technical University May 15, 2023 Abstract Retinal optical coherence tomography (OCT) and optical coherence tomography angiography (OCTA) are promising tools for the early-stage diagnosis of Alzheimer’s disease (AD). These non-invasive imaging techniques are cost-effective and more accessible than alternative neuroimaging tools. However, the current literature lacks an extensive review of AD or cognitive impairment diagnosis using OCT or OCTA. This motivated us to examine recent deep learning studies using the PRISMA approach to systematic review. We used Publish or Perish software to locate relevant research from databases such as Scopus, PubMed, and Web of Science, obtaining an initial pool of 2725 references. We then followed the PRISMA review process to identify twelve relevant studies and two patent applications for detailed analysis. Half of the papers we reviewed described longitudinal mouse studies targeting early AD detection. Whereas earlier research used patient demographics and pre-computed features as inputs to the classical machine learning classifiers, more recent studies tend to employ end-to-end deep learning models with OCT and OCTA image inputs. However, this approach presents issues such as small datasets and an absence of scan- ning standards, which the reviewed literature addresses in various ways. We discuss the lack of open OCT/OCTA datasets (about Alzheimer’s disease) as the main issue impeding progress in the field. DOI: xxx/xxxxARTICLE TYPE Automated diagnosis of AD using OCT and OCTA: A systematic review Yasemin Turkan* 1 | Faik Boray Tek 2 1Department of Computer Engineering, Isik University, Istanbul, Turkey 2Department of Artificial Intelligence and Data Engineering Istanbul Technical University, Istanbul, TurkeyCorrespondence*Yasemin Turkan, Mesrutiyet Koyu, Universite Sok. No:2, Sile, 34980, ˙ Istanbul, Turkey Email: turkan.yasemin@gmail.comPresent AddressPresent address INTRODUCTION Alzheimer’s disease (AD) is an irreversible and progressive brain disorder characterized by a decline in cognitive function, and is the most prevalent type of dementia. It currently has no known cure and is marked by a significant reduction in brain size (neurodegeneration) caused by the accumulation of proteins (amyloid-beta and tau) in neurons. 1 As the retina and brain originate from the same neural tube, the eyes are often regarded as an extension of the brain, 2 and postmortem studies show that amyloid- beta and tau protein accumulate in the retinas of individuals with AD. 3 High-resolution visual imaging technologies, such as optical coherence tomography (OCT) and optical coherence tomography angiography (OCTA) have recently been proposed to examine structural and vascular changes in the retinas of AD patients. In a recent study, Vij et al. 4 reviewed advances in retinal modalities for diagnosing Alzheimer’s disease. Their review 1