This study investigates the use of Artificial Intelligence (AI) and Deep Learning algorithms to detect and analyze hypertensive retinal changes through fundus photography, with a focus on integrating genetic data, clinical characteristics, and medical history to improve predictive models.
Hypertension is a leading cause of retinal vascular damage, potentially resulting in irreversible vision loss. Current diagnosis and monitoring rely on inaccurate and stressful blood pressure measurements, leading to false positives and negatives. The retina offers a unique window into the vascular effects of hypertension, but early signs of hypertensive retinopathy are difficult to detect, even for experienced ophthalmologists.
Recent advances in AI and Deep Learning have shown promise in analyzing medical images, including retinal fundus photographs. To develop a comprehensive predictive model, we will incorporate clinical characteristics, such as blood pressure readings, diabetes diagnoses, and medical history, alongside genetic data. By integrating these multi-modal data sources, we aim to develop more accurate and personalized screening tools for hypertensive retinopathy.
Objectives:
1. Assess the ability of AI-driven deep learning algorithms to detect hypertensive retinal changes from fundus photographs.
2. Evaluate the accuracy, sensitivity, and specificity of AI models in identifying hypertensive retinopathy, incorporating genetic data, clinical characteristics, and medical history.
3. Develop an accessible, scalable, and cost-effective telemedicine solution for hypertensive retinopathy screenings, leveraging the integration of AI, retinal imaging, genetic data, and clinical information.
