Emerging research suggests that retinal biomarkers can be used to diagnose Alzheimer’s disease in its early stages. This is based on the observation that changes in the structure of the retina can occur before the onset of clinical symptoms. While experienced ophthalmologists have been able to identify these changes for years, it has been difficult to develop definitive criteria for the codification of the specific patterns. Machine learning offers a promising approach to overcoming this challenge.
In addition to machine learning, adaptive optics and holographic-based approaches can be used to significantly boost the resolution of captured images. This is done by dynamically correcting for imperfections in the lenses, allowing for resolutions that approach the theoretical limit imposed by classical physics. The author has made contributions to this area of research by developing the algorithms and modulation hardware to effect these improvements (Kadis, 2020, 2021, 2022).
This research proposal seeks to combine these two aspects to improve the efficacy of early-stage detection of Alzheimer’s disease. However, collecting and labelling data over a 20-year period to build a comprehensive early-detection pipeline is not practical. Therefore, the authors propose to use the UK Biobank and to emulate the boosted optical imaging performance using simulation.
We will use state-of-the-art image upscaling techniques (based on deep learning) to emulate adaptive-optics fundus imaging system. We shall then build two separate AI/machine learning pipelines, one with the standard biobank dataset and one with the AO-emulated dataset. This will allow us to understand if there is any scientific justification for using boosted resolution images using AO when looking for early-detection retinal biomarkers for Alzheimer’s disease.
The success of this research could have a significant impact on the diagnosis and treatment of Alzheimer’s disease. By enabling early detection, clinicians may be able to intervene before irreversible damage occurs, potentially slowing or even halting the progression of the disease.