The main purpose of this study is to pioneer an easy risk stratification tools, which is developed using novel artificial intelligence (AI) algorithms, that will be able to detect common and fatal heart diseases easily simply through a picture of the back of the eye, the retina. The retinal images will be analysed using a computer application with the risk stratification tool to predict health outcome of individual. The study also aims to correlate between clinical characteristics, lifestyle (eg. exercise, sleep, erectile dysfunction) and diet to retinal and coronary vasculature and clinical outcomes.
Cardiovascular disease (CVD) is ranked 1st for mortality rate globally. In 2016, approximately 17.9 million people died from CVDs, representing 31% of all global deaths. Retinal vasculature has been characterised as the 'window' to the body's circulatory system and been correlated with several diseases that perturbate systemic micro- and macro-vasculature such as hypertension, stroke and chronic kidney disease. As microvascular changes often precede macrovascular changes, retinal imaging has the capability to be an easily available, non-invasive biomarker to screen for multiple vascular pathologies. A deep learning system (DLS) is a novel,state-of-art artificial intelligence (AI) technology that has achieved robust diagnostic performance for medical imaging analysis. The integration of deep learning (DL) into retinal image evaluation has accelerated its potential further. Its examination of fundus photographs has demonstrated abilities to detect several signs that are undetected by the human eye. With recent correlations being established between CAD and retinal vasculature aberrancies such as reduced calibres and fractal dimensions, there is a clear niche for AI to predict cardiac diseases with the use of retinal fundus photographs.
Study Type
OBSERVATIONAL
Enrollment
2,000
The retinal photography will be acquired from macula-centre and an optic-disc centre for both left and right eye after their cardiac assessment. Patient will be required to focus at a target inside the machine for approximately 5 to 10 seconds, with a flash followed at the end when eye image is captured. Based on patients' eye conditions, additional or less images may be taken. This allows a snapshot of both retinal and coronary vessels at a single point in time, no mydriatic drops will be used. Participants will be given the option to undergo mydriatic photographs of the retina at SNEC/SERI depending on participant eye condition. For mydriatic photograph, images will be obtained after pupil dilation and graded for presence and severity of diabetic retinopathy according to the modified ETDRS system and other retinal abnormalities by trained graders.
3 to 4 questionnaires will be administered (i.e. EQ5D, International Index of Erectile Function, Pittsburgh Sleep Quality Index, General Questionnaire) to investigate the correlation between clinical characteristics, lifestyle (eg. exercise, sleep, erectile dysfunction) and diet to retinal and coronary vasculature and outcomes.Only male patients who are literate in English language will be invited to complete the International Index of Erectile Function (IIEF) questionnaire. Participants have the option to not complete the whole IIEF questionnaire or some of the questions if they are not comfortable with answering those questions.
National Heart Centre Singapore
Singapore, Singapore
AI model and retinal imaging
This study will also expand the potential for AI models and retinal imaging as tools to identify systemic micro and macrovascular diseases such as hypertension, stroke and provide an easily available, risk-free tool for physicians to readily evaluate the state of the body's vessels.
Time frame: Through study completion, an average of 2 years
cross sectional performance of Cardiovascular risk stratification tool
Evaluate the cross sectional performance of Cardiovascular risk stratification tool, SIVA-DLS and Retina-CAC for cardiovascular risk stratification compared with patients undergoing cardiac CT scan. Low risk is determined to be Agatston calcium score=0 and high risk being Agatston calcium score 100 and above.
Time frame: Through study completion, an average of 2 years
longitudinal major cardiovascular event rate in subjects identified as high risk
Evaluate the longitudinal major cardiovascular event rate in subjects identified as high risk using a combination of cardiovascular risk stratification tool, SIVA DLS and Retina-CAC
Time frame: Through study completion, an average of 2 years
major cardiovascular event rate in subjects identified as high risk
Evaluate if the major cardiovascular event rate in subjects identified as high risk and placed under medical intervention is lower compared to matched historical controls.
Time frame: Through study completion, an average of 2 years
vision threatening retinopathy(VTDR) rate and rate of vision loss in subjects identified as high risk for VTDR
Evaluate if the vision threatening retinopathy(VTDR) rate and rate of vision loss in subjects identified as high risk for VTDR and placed in a VTDR clinic is lower compare to matched historical controls.
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.
Time frame: Through study completion, an average of 2 years