| Title: | Mobile Application for Alzheimer's disease Diagnosis Using Deep-Learning. |
| Lead Institution: | Ajou University |
| Principal investigator: | Mr Doyoung Kim |
Application 73548
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
About
This research aims to develop a Deep-Learning (a type of machine learning considered to be in some way more dynamic or complete than others.) model for Alzheimer's disease diagnosis whose goal is to achieve accuracy at least up to 90%, and implement the model on a smartphone application. Some studies show retinal related disorders in the human neural system. Specifically our research focuses on studies about retinal biomarkers (a diagnostic indicator of a medical condition.) for Alzheimer's disease. These studies found abnormalities in the retina using OCT(Optical Coherence Tomography) or retinal photograph. Alzheimer's disease is the leading cause of dementia. The number of Americans age 65 and older with Alzheimer's dementia may grow to 13.8 million by 2050. This represents a steep increase from the estimated 5.8 million Americans age 65 and older who have Alzheimer's dementia today. There is a study about the Deep-Learning model for Alzheimer's disease classification. They showed the possibility of using Deep-learning models for diagnosing Alzheimer's disease. In their study, the model achieved 82.44% accuracy. We aim to finish and get results in this research for 12 months. We classify datasets from UK Biobank according to retina image quality for training. We will also develop an algorithm that can classify fundus retina images according to image quality. This algorithm avoids human error, and consequently aids in model accuracy. Then our model is trained with a qualified image dataset. This model may achieve at least 90% accuracy for binary classification in Alzheimer's disease diagnosis. After training, the model will be developed into a small and lightweight model that will be implemented as a smartphone app. Finally, our research team will use a lens and smartphone to get a fundus retinal images dataset and test our smartphone app., which will operate based on datasets from your institute. In the case of Alzheimer's disease, it is important to detect it at early stages, because the treatment we receive now is to slow the progression of the disease. Therefore, we expect individuals to save money for Alzheimer's treatment after the smartphone application is developed in our study. In addition, these results will reduce the cost of social maintenance and alleviate the national disease burden. The algorithms in our study can be used for other medical imaging diagnostics. In particular, detection of other diseases on fundus retinal photographic images.1 Publication
| Pub ID | Title | Author(s) | Year | Journal |
|---|---|---|---|---|
| 12194 | Efficient Deep Retinal Fundus Image-Based Network for Alzheimer's Disease Diagnosis Using Mobile Device Applications | Young Kim (+3) | 2024 | IEEE Access |
Enabling scientific discoveries that improve human health