Predicting High Coronary Artery Calcium Score From Retinal Fundus Images With Deep Learning Algorithms

Abstract

To evaluate high accumulation of coronary artery calcium (CAC) from retinal fundus images with deep learning technologies as an inexpensive and radiation-free screening method. Methods: Individuals who underwent bilateral retinal fundus imaging and CAC score (CACS) evaluation from coronary computed tomography scans on the same day were identified. With this database, performances of deep learning algorithms (inception-v3) to distinguish high CACS from CACS of 0 were evaluated at various thresholds for high CACS. Vessel-inpainted and fovea-inpainted images were also used as input to investigate areas of interest in determining CACS. Results: A total of 44,184 images from 20,130 individuals were included. A deep learning algorithm for discrimination of no CAC from CACS >100 achieved area under receiver operating curve (AUROC) of 82.3% (79.5%-85.0%) and 83.2% (80.2%-86.3%) using unilateral and bilateral fundus images, respectively, under a 5-fold cross validation setting. AUROC increased as the criterion for high CACS was increased, showing a plateau at 100 and losing significant improvement thereafter. AUROC decreased when fovea was inpainted and decreased further when vessels were inpainted, whereas AUROC increased when bilateral images were used as input. Conclusions: Visual patterns of retinal fundus images in subjects with CACS > 100 could be recognized by deep learning algorithms compared with those with no CAC. Exploiting bilateral images improves discrimination performance, and ablation studies removing retinal vasculature or fovea suggest that recognizable patterns reside mainly in these areas. Translational Relevance: Retinal fundus images can be used by deep learning algorithms for prediction of high CACS.

Retinal Microvascular Changes in Familial Hypercholesterolemia: Analysis with Swept-Source Optical Coherence Tomography Angiography

“…Recently, a large study highlighted the value of retinal photograph-based deep learning as an alternative measure of CAC score and for the prediction of cardiovascular events [ 37 , 38 ]. However, to our knowledge, there is no study to date analyzing OCT-A-derived retinal microvascular measures and the CAC score.…”

Deep-learning-based cardiovascular risk stratification using coronary artery calcium scores predicted from retinal photographs

“…The novelty of the present study is further extending the previous work by Poplin and colleagues 12 to cardiac CTmeasured CAC and its value in predicting future cardiovascular disease events and mortality. Recently, a deeplearning algorithm to predict abnormal CAC from retinal photographs was reported, 26 but the algorithm was not validated in independent datasets, and its ability to predict future cardiovascular disease events was not evaluated.…”

Artificial Intelligence in Predicting Systemic Parameters and Diseases From Ophthalmic Imaging

“…Cardiovascular parameters predicted from RFP include systolic and diastolic blood pressure (BP), hypertension, retinal vessel caliber, coronary artery calcium (CAC) and carotid artery atherosclerosis ( 21 , 24 , 36 – 38 , 41 , 46 , 47 , 69 ). RFP are thought to be robust input images for predicting cardiovascular disease, as they can directly capture many retinal features associated with increased cardiovascular risk, including vessel caliber, tortuosity, and bifurcations ( 70 , 71 ).…”

“…CAC scores of 401–999 had RR of 7.2 (95% CI 5.2–9.9), and CAC score of 1,000 had RR of 10.8 (95% CI 4.2–27.7) ( 73 ). Son et al ( 41 ) predicted abnormal CAC scores at various thresholds, producing an AUC of 0.832 when the threshold was set at >100 units. Furthermore, Rim et al ( 37 ) derived a deep learning-based CAC score predicted from RFP (RetiCAC) and used this new RetiCAC score for cardiovascular risk stratification.…”