High-Throughput, Contact-Free Detection of Atrial Fibrillation From Video With Deep Learning.

This study uses video and a pretrained deep convolutional neural network to analyze facial photoplethysmographic signals in detection of atrial fibrillation.