Researchers for an interesting new paper suggested that a new algorithm may make it possible to assist in the diagnosis of coronary artery disease (CAD) with a facial photograph.

The paper, published in the European Journal of Cardiology, was a multicenter, cross-sectional study of patients undergoing coronary angiography or CT angiography at nine sites in China. The purpose of evaluating the scans was to train and validate a deep convolutional neural network for CAD detection (at least one ≥50% stenosis) from facial photographs. The analysis included 5,796 consecutively enrolled patients who were randomly assigned to either training (n=5,216) or validation (n=580) groups for the development of the algorithm. They then enrolled 1,013 patients into the algorithm test group and calculated sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) using radiology-based diagnosis as the standard.

According to the results, using the operating cut point with high sensitivity, the detection algorithm had a sensitivity of 0.80 and a specificity of 0.54 in the test cohort (AUC was 0.730; 95% CI, 0.699 to 0.761). They also reported that the algorithm AUC was higher for the Diamond-Forrester model (0.730 vs. 0.623; P<0.001) as well as the CAD consortium clinical score (0.730 vs. 0.652; P<0.001).

“To our knowledge, this is the first work demonstrating that artificial intelligence can be used to analyse faces to detect heart disease,” lead author Professor Zhe Zheng, vice director of the National Center for Cardiovascular Diseases and vice president of Fuwai Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, People’s Republic of China, said in a press release. “It is a step towards the development of a deep learning-based tool that could be used to assess the risk of heart disease, either in outpatient clinics or by means of patients taking ‘selfies’ to perform their own screening. This could guide further diagnostic testing or a clinical visit.”

Credit: Original article published here.